Shared tray leasing scheme optimization method and system

Abstract

The invention relates to a method and a system for optimizing a leasing scheme of a shared tray, belongs to the technical field of logistics equipment sharing, and discloses the method and the system for optimizing the leasing scheme of the shared tray by acquiring historical leasing data, real-time inventory data and predicted demand data of each node of a supply chain; pallet full life cycle trajectory data is acquired based on an RFID technology, a pallet circulation digital twin model is established, a full process state is mapped, multi-objective optimization is performed on a pallet scheduling path, inventory distribution and lease pricing through an improved genetic algorithm, and intellectualization, high efficiency and precision of a shared pallet lease full process are realized. The problems that in the prior art, scheduling is low in efficiency, pricing is rigid, a management closed loop is missing, and multi-terminal collaboration is insufficient are solved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of logistics equipment sharing, and particularly relates to a shared pallet rental scheme optimization method and system. BACKGROUND

[0002] In supply chain management, pallets as key logistics carrying equipment, their recycling efficiency directly affects logistics cost and management level. The existing shared pallet rental scheme has many deficiencies: first, pallet scheduling relies on manual planning, and it is difficult to dynamically optimize according to real-time inventory, network distribution and order demand, resulting in coexistence of idle resources and supply and demand imbalance; second, the rental pricing adopts fixed rules, and does not accurately calculate multi-dimensional factors such as customer credit rating, rental period, and pallet wear rate, affecting platform revenue and customer satisfaction; third, the pallet life cycle management lacks a closed loop, and the data of rental, use, recycling, and maintenance are fragmented, and real-time early warning and efficient handling of abnormal situations cannot be achieved; fourth, there is information delay in multi-end collaboration (such as platform end, network end, customer end and APP end), resulting in slow order response and poor process connection.

[0003] Although the shared pallet system based on RFID technology has realized basic information management, the existing technology does not fully exploit the value of data, and the optimization mechanism is missing, making it difficult to meet the needs of efficient operation of the supply chain. SUMMARY

[0004] The present application provides a shared pallet rental scheme optimization method and system, which is used to solve the problems of inefficient scheduling, rigid pricing, missing management closed loop and insufficient multi-end collaboration in the prior art, and realizes intelligent, efficient and accurate of the whole process of shared pallet rental by improving the genetic algorithm for multi-objective optimization of pallet scheduling path, inventory distribution and rental pricing.

[0005] In order to achieve the above purpose, the present application realizes the following technical scheme: A shared pallet rental scheme optimization method, comprising the following steps: Obtain historical rental data, real-time inventory data and predicted demand data of each node of the supply chain; Collect pallet life cycle trajectory data based on RFID technology, establish a pallet flow digital twin model, and map the whole process state of the pallet from warehousing, rental, transportation, recycling to maintenance / scraping; Multi-objective optimization of pallet scheduling path, inventory distribution and rental pricing is performed by improved genetic algorithm.

[0006] Further, in the step of obtaining data of each node of the supply chain, an edge computing node is deployed at the warehouse / network point end to collect pallet in-out warehouse data, equipment running state data and environmental temperature and humidity data in real time; The pre-processed data is uploaded to the cloud platform through the 5G communication module. The pre-processing process includes data deduplication, outlier removal, and format standardization. Outlier removal is based on the 3σ principle to identify abnormal loss rate and inventory quantity data.

[0007] Further, in the step of establishing the tray circulation digital twin model, the unique identification of the tray collected by the RFID is bound to the virtual tray in the digital twin model, and the physical state parameters of the tray are synchronized in real time; The physical state parameters include wear degree, bearing history, location coordinates, and associated order information. The remaining service life of the tray is predicted through time series analysis, and when the remaining service life is lower than the first preset threshold, a maintenance reminder or scrap warning is automatically triggered.

[0008] Further, in the improved genetic algorithm optimization step, the coding method uses real number coding, and the tray scheduling path, inventory allocation ratio, and pricing coefficient are used as gene segments. The fitness function is constructed as transportation cost weight coefficient x transportation cost value + inventory turnover rate weight coefficient x (1-inventory turnover period) + customer satisfaction weight coefficient x customer satisfaction value. The weight coefficients are determined by the analytic hierarchy process. In the genetic operation, adaptive crossover probability and mutation probability are introduced, which are dynamically adjusted according to the evolution generation of the population, to avoid the algorithm falling into local optimal solution.

[0009] Further, it includes a lease pricing dynamic adjustment step: a pricing factor system is constructed based on customer credit rating, lease batch size, lease period, and historical cooperation frequency. Customer credit rating is evaluated comprehensively based on account performance record, settlement timeliness, and tray loss rate. A gradient boosting regression model is used to train the pricing factors to generate an individualized pricing plan. When the customer lease batch size exceeds the second preset threshold or the lease period covers the industry off-season, the discount coefficient is automatically adjusted.

[0010] Further, it includes an order intelligent matching step: based on the tray type, quantity, delivery address, and usage period requirement in the customer order, combined with the optimized inventory distribution data, the optimal outbound warehouse / point and carrier are automatically matched. In the matching process, the warehouse / point closest to the customer and with the highest available tray rate is preferred, and the historical on-time delivery rate and transportation cost offer of the carrier are also considered.

[0011] Further, it includes a recycling optimization step: based on the recycling application data initiated by the member end, the geographical distribution of the point, and the tray state data, the optimal recycling route is planned, and the clustering algorithm is used to combine adjacent recycling demands to reduce the empty mileage of the recycling vehicle. The recovered tray is classified and detected, and is distributed to a reuse, repair or scrap channel according to the detection result, and the detection indexes include appearance integrity, structural strength and RFID signal stability.

[0012] Further, the abnormal order intelligent processing step is included: the lease order state is monitored in real time through a machine learning model, and abnormal conditions such as order delay, tray damage and settlement exception are identified; The machine learning model adopts a random forest algorithm and is generated based on historical abnormal order data training; Different types of abnormal triggers trigger corresponding processing strategies, automatically adjust the subsequent scheduling plan when the order is delayed, start the claim process and supplement the replacement tray when the tray is damaged, and push the reminder information to the financial module when the settlement is abnormal.

[0013] Further, the optimization effect evaluation step is included: an evaluation index system is established, and the index system includes cost reduction rate, inventory turnover improvement rate, order response time shortening rate, customer complaint rate and tray recycling rate; The optimized actual operation data and the benchmark data before optimization are compared and analyzed regularly, the parameters and optimization target weights of the genetic algorithm are dynamically adjusted according to the evaluation result, and the continuous iterative optimization of the lease scheme is realized.

[0014] A shared tray lease scheme optimization system for executing the above method, comprising: The data acquisition and preprocessing module is used for acquiring historical lease data, real-time inventory data and predicted demand data of each node of the supply chain; The digital twin modeling module establishes a tray circulation digital twin model based on the tray full-life cycle track data collected by the RFID technology, binds the tray unique identifier collected by the RFID with the virtual tray in the digital twin model, synchronizes the physical state parameters of the tray in real time and maps the full-process state; The multi-objective optimization module adopts an improved genetic algorithm to perform multi-objective optimization on the tray scheduling path, inventory distribution and lease pricing; And the function execution module includes a lease pricing dynamic adjustment unit, an order intelligent matching unit, a recovery optimization unit, an abnormal order intelligent processing unit and an optimization effect evaluation unit; the function execution module is configured to execute the steps of the above method.

[0015] As can be seen from the above, the tray circulation digital twin model is established to track the tray state in real time, and the improved genetic algorithm is used to realize multi-objective dynamic optimization, which solves the problems of low resource scheduling efficiency, rigid pricing mechanism and lack of full-process management in the prior art, and has the significant advantages of improving supply chain coordination efficiency and reducing operation cost. BRIEF DESCRIPTION OF DRAWINGS

[0016] The accompanying drawings, which are incorporated herein and constitute part of the specification, illustrate embodiments consistent with the application and, together with the description, function to explain the principles of the application.

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings required by the embodiments or the prior art description will be briefly introduced as follows. Obviously, for those skilled in the field, other drawings can also be obtained based on these drawings without any creative work.

[0018] Figure 1 A workflow diagram of the present application; Figure 2 A system internal structure diagram of the present application; Figure 3 A functional execution module internal structure diagram of the present application.

[0019] Figure legend: 10, data acquisition and preprocessing module; 20, digital twin modeling module; 30, multi-objective optimization module; 40, function execution module; 401, lease pricing dynamic adjustment unit; 402, order intelligent matching unit; 403, recycling optimization unit; 404, abnormal order intelligent processing unit; 405, optimization effect evaluation unit.

[0020] The purpose of the present application, functional characteristics and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0021] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and not to limit the present application. In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below in conjunction with the drawings and specific embodiments.

[0022] For the shared pallet leasing scheme in supply chain management, there are problems such as low resource scheduling efficiency, inflexible pricing mechanism, non-closed-loop whole life cycle management, and delayed multi-end collaborative information. The traditional method relies on manual planning and scheduling path, which is difficult to dynamically adjust according to real-time inventory and order demand, resulting in coexistence of pallet idle and shortage. The fixed rule is used for lease pricing, which does not combine customer credit rating and pallet wear rate for dynamic adjustment, affecting the operation income. The data of each link in the pallet circulation process is isolated, and the whole life cycle state tracking cannot be realized, and the abnormal situation is difficult to handle in time.

[0023] To solve the above problems, the prior art data collection dimension is single, lacking real-time monitoring of the physical state of the pallet, leading to lagging scheduling decisions. Through analysis, it is found that establishing a full-process data chain is the key foundation for dynamic optimization. To solve the problem of multi-objective collaborative optimization, genetic algorithm is introduced into the scheduling model, but the traditional algorithm is easy to fall into local optimum. After many experimental verifications, using real number coding combined with adaptive parameter adjustment can effectively improve the convergence of the algorithm. To realize closed-loop management, it is proposed to bind the physical parameters collected by RFID with the digital twin model and update the pallet state synchronously.

[0024] Therefore, as Figure 1 shown, the present embodiment proposes a shared pallet rental scheme optimization method, including the following steps: Step S100: Obtain historical rental data, real-time inventory data and predicted demand data of each node of the supply chain; Step S200: Collect pallet full-life cycle trajectory data based on RFID technology, establish a pallet flow digital twin model, and map the full-process state of the pallet from warehousing, rental, transportation, recycling to maintenance / scraping; Step S300: Multi-objective optimization of pallet scheduling path, inventory distribution and rental pricing is performed by improved genetic algorithm.

[0025] Among them, the historical rental data refers to the pallet rental records of each node in the past period, which is specifically obtained through a database query interface, and is used to analyze customer rental behavior patterns. Real-time inventory data refers to the number of pallets currently available in the warehouse or network, which can be specifically collected in real time by Internet of Things sensors to provide a basis for dynamic scheduling. The predicted demand data refers to the pallet demand in each region in the future time period, which can be specifically generated by using a time series prediction model, and is used to predict resource gaps. The full-life cycle trajectory data refers to the state change records generated during the flow process of the pallet, which can be specifically collected by RFID readers at key nodes to record position, timestamp and operation type, forming a complete flow chain. The digital twin model refers to a dynamic mapping model of the pallet constructed in a virtual space, which can be specifically implemented by using three-dimensional modeling technology to realize real-time synchronization of physical parameters and virtual models, supporting full-process state visualization. The improved genetic algorithm refers to the optimized swarm intelligence algorithm, which can be specifically expressed by using real number coding to represent the scheduling path and inventory allocation scheme, and by using adaptive crossover probability and mutation probability to improve the global search ability.

[0026] Specifically, the historical data of each node in the supply chain is transmitted to the central processing platform through the data interface to form a multidimensional data set. The RFID reader automatically scans the electronic tag when the pallet enters or exits the warehouse, records the time, location and operation type, and generates full life cycle tracking data. The digital twin model updates the state parameters of the virtual pallet after receiving real-time data, such as updating the location coordinates and transportation time when the pallet is transported to the transfer station. The improved genetic algorithm encodes the scheduling path as a real number sequence, with each gene segment corresponding to the selection order of the warehouse node, and the inventory allocation ratio is encoded as the resource proportion of each node. During the iteration process of the algorithm, the fitness function calculates the weighted sum of transportation cost, inventory turnover rate and customer satisfaction, and selects high-quality individuals through the roulette strategy. The two-point crossover method is used for crossover operation to exchange parent gene segments, and the mutation operation randomly adjusts part of the gene values to increase population diversity.

[0027] Compared with the prior art, the traditional scheme relies on manual experience to develop a scheduling plan, which is difficult to cope with dynamic changes in demand fluctuations. The present scheme realizes multi-objective automatic optimization through genetic algorithm, significantly improving decision-making efficiency. In the prior art, pallet state data is scattered in independent systems and cannot form a full-process management closed loop. The present scheme realizes real-time mapping between the physical world and the virtual space through the digital twin model, so that abnormal states can be identified in a timely manner. The traditional pricing model uses fixed coefficients to calculate rental fees. The present scheme dynamically adjusts the pricing strategy in combination with real-time inventory data and predicted demand, enhancing market adaptability.

[0028] Through the above technical means, the embodiment can dynamically optimize the pallet scheduling path, reduce the empty load rate of transportation vehicles, realize reasonable allocation of inventory resources, reduce the risk of regional shortages, establish a full life cycle data chain, and improve the tracking accuracy of pallet state. Through the multi-objective collaborative optimization mechanism, the cost control and customer satisfaction are balanced. The rental pricing strategy can be adjusted flexibly according to market demand, enhancing the platform's revenue capacity. The digital twin model provides data support for abnormal early warning, shortening the fault response time.

[0029] Further, in the data collection step at each node of the supply chain, an edge computing node is deployed at the warehouse or network point to collect pallet in-out data, equipment running state data and environmental temperature and humidity data in real time. The preprocessed data is uploaded to the cloud platform through the 5G communication module. The preprocessing process includes data deduplication, outlier elimination and format standardization. Outlier elimination identifies loss rate and inventory quantity data that exceeds the normal range based on the 3σ principle.

[0030] The edge computing node refers to a distributed computing device deployed in a physical warehouse or service point, which can be implemented by an embedded industrial computer combined with an Internet of Things gateway, and is used for processing local sensor data on site to reduce transmission delay. The 5G communication module refers to a wireless transmission device supporting the fifth generation mobile communication technology, which can be implemented by an industrial-grade communication module integrated with a 5G baseband to ensure low-latency transmission of high-concurrency data. Data deduplication refers to the process of eliminating duplicate collection or redundant records, which can be implemented by a hash value comparison algorithm to reduce invalid data storage resources. Outlier rejection refers to filtering abnormal data that deviates from the reasonable range, which can be implemented by calculating the data fluctuation interval based on the 3σ principle of normal distribution to automatically filter out abnormal records that deviate from the mean value by more than three times the standard deviation. Format standardization refers to unifying the storage structure of multi-source heterogeneous data, which can be implemented by a JSON format conversion tool to convert raw data collected by different devices into standardized data packets with uniform fields.

[0031] Specifically, the edge computing node directly accesses the RFID reader, temperature and humidity sensor, and warehouse management system in the warehouse to obtain real-time data such as pallet in-out timestamp, forklift running status, and warehouse environment parameters with millisecond-level response speed. During the collection process, the data deduplication module eliminates redundant information caused by repeated reporting of sensors by comparing the hash values of adjacent data records in the time series. The outlier rejection module performs statistical distribution analysis on data with sudden changes in inventory quantity or abnormal increase in pallet loss rate, dynamically calculates the normal value interval using the 3σ principle, and automatically filters out abnormal data points that exceed the threshold range. The data packets processed by format standardization are transmitted to the cloud through the 5G communication module, with a transmission delay of less than 50 milliseconds, ensuring real-time synchronization of supply chain dynamic data.

[0032] Compared with the prior art, the traditional supply chain data collection relies on the central server to poll each node data, which has a minute-level delay and is easily affected by network fluctuations. This scheme realizes data processing on site through the edge computing node, shortens the collection response time from minutes to milliseconds, and uses 5G communication to ensure stable transmission of massive data. The existing technology mainly uses manual sampling method to process abnormal data, and this scheme realizes automatic identification and filtering of abnormal values through the 3σ algorithm, which improves the data cleaning efficiency by two orders of magnitude and effectively avoids the interference of false data on subsequent optimization algorithms.

[0033] Through this technical means, the core problems of insufficient real-time data collection and low data quality of supply chain nodes are solved. The distributed deployment of edge computing nodes reduces the data collection delay to a level that the business system can respond in real time, the 5G communication module guarantees the efficient transmission of preprocessed data, the statistical-based outlier removal mechanism significantly improves the accuracy of inventory and loss data, and the format standardization processing eliminates the structural barriers of multi-source data fusion. These technical means together build a high-quality data collection system, providing reliable data support for pallet scheduling path optimization and dynamic inventory allocation.

[0034] Further, the embodiment proposes that in the step of establishing the pallet flow digital twin model, the RFID-acquired unique identification of the pallet is bound to the virtual pallet in the digital twin model, and the physical state parameters of the pallet are synchronized in real time; the physical state parameters include wear degree, bearing history, position coordinates and associated order information, the remaining service life of the pallet is predicted through time series analysis, and when the remaining service life is lower than the preset threshold, a maintenance reminder or a scrap warning is automatically triggered.

[0035] Among them, the RFID unique identification refers to the identity code assigned to each physical pallet through radio frequency identification technology, which can be implemented by a passive electronic tag conforming to the ISO 18000-6C standard, and is used to automatically identify and track individual pallets in the supply chain. The digital twin model refers to a virtualized pallet object constructed by three-dimensional modeling and Internet of Things data fusion, which can be implemented by Unity3D engine and MQTT protocol, and is used to map the state changes of the physical pallet in real time. The wear degree in the physical state parameters refers to the fatigue damage value of the pallet surface and structural parts, which can be obtained by collecting deformation data with a laser scanner and calculating by finite element analysis, and the bearing history refers to the load record of the pallet in each rental task, which can be obtained by associating pressure sensors and order databases. Time series analysis refers to a method for dynamically predicting the performance parameters of the pallet, which can be implemented by ARIMA model combined with sliding window mechanism, and is used to capture the performance degradation law of the pallet.

[0036] Specifically, when the pallet enters the warehouse, the RFID reader automatically reads its unique identification code and generates a corresponding virtual pallet instance in the digital twin system. The virtual model receives real-time data from the pallet sensors through the Internet of Things gateway, including vibration frequency, load weight, and environmental temperature and humidity parameters during each transportation task. The load history data is stored as a time-stamped sequence, associated with the type of goods in the order system and the transportation mileage information. The degree of wear is calculated by periodically scanning the pallet structure and comparing it with the initial model to form a time series of material fatigue coefficients. These parameters are input into the trained ARIMA prediction model, which generates a prediction curve of the remaining service life by analyzing the autocorrelation and difference stationarity of historical data. When the predicted value is below the preset maintenance threshold, the system automatically generates a work order and assigns it to the nearest maintenance site; if the predicted value is below the scrap threshold, the replacement process is triggered and the inventory data is updated.

[0037] Compared with the prior art, the traditional pallet management relies on manual inspection and regular maintenance plans, which cannot real-time perceive the state change of the pallet, resulting in delayed maintenance response or over-maintenance. Although RFID is used for pallet tracking in the prior art, a virtual-real mapping twin model is not established, and the physical parameter collection is limited to location information, lacking dynamic monitoring of structural performance. The present scheme constructs a full-dimensional state evaluation system by fusing multi-source sensor data and business system information, and uses time series analysis to replace the fixed cycle detection mechanism, realizing accurate prediction and hierarchical processing of maintenance needs.

[0038] Through the above technical means, the present application realizes real-time synchronization of the physical state of the pallet and the digital model, solving the data lag problem in traditional management. Time series analysis based on dynamic parameters can accurately capture the performance degradation trend of the pallet, avoiding false judgments caused by fixed threshold judgments. The dual triggering mechanism of maintenance reminders and scrap warnings can take differentiated handling measures according to the actual degree of wear, prolonging the service period of usable pallets and preventing safety hazards. The load history record associated with order information provides a data basis for analyzing pallet usage patterns, which helps to optimize subsequent scheduling strategies and inventory distribution.

[0039] Further, the present embodiment proposes that in the optimization step of the improved genetic algorithm, the coding method uses real number coding, and the pallet scheduling path, inventory allocation ratio, and pricing coefficient are taken as gene segments; the fitness function is constructed as transportation cost weight coefficient x transportation cost value + inventory turnover rate weight coefficient x (1-inventory turnover period) + customer satisfaction weight coefficient x customer satisfaction value, and the weight coefficients are determined by the analytic hierarchy process; adaptive crossover probability and mutation probability are introduced in the genetic operation, which are dynamically adjusted according to the population evolution generation number, avoiding the algorithm falling into local optimal solution.

[0040] The real number coding refers to directly taking the coordinate sequence of the scheduling path, the inventory allocation proportion value and the pricing coefficient value as the chromosome gene, and can be specifically realized by using a continuous variable coding mode, so that the multi-dimensional decision variable is continuously distributed in the solution space. The compound fitness function refers to linearly weighting the three optimization targets of transportation cost, inventory turnover rate and customer satisfaction, and the weight proportion of each target can be specifically determined by using the analytic hierarchy process, for example, the weight coefficient is calculated by comparing the importance of each target by a domain expert. The adaptive genetic operation refers to dynamically changing the crossover probability and mutation probability with the increase of the evolution generation, and the crossover probability can be specifically adjusted by using an exponential decay function, and the mutation probability can be dynamically adjusted according to the population diversity index.

[0041] Specifically, in the initialization stage of the genetic algorithm, the latitude and longitude coordinate sequence of the scheduling path, the inventory allocation percentage value of each warehouse and the pricing coefficient value of different customer groups are coded into real number type gene fragments. In the iteration process, the transportation cost value is calculated by the path distance and the unit transportation cost, the inventory turnover period is determined by the ratio of the inventory quantity to the daily consumption, and the customer satisfaction value is calculated by the comprehensive score of the historical order fulfillment rate and the complaint rate. The analytic hierarchy process allocates weights to the three optimization targets according to business requirements, for example, the transportation cost weight is set to 0.5, the inventory turnover rate weight is set to 0.3, and the customer satisfaction weight is set to 0.2. With the increase of the evolution generation, the crossover probability is gradually reduced from the initial value 0.8 to 0.4 according to the exponential function, and the mutation probability is gradually increased from 0.01 to 0.1 according to the similarity index of the population individuals, so as to balance the global search and local optimization ability.

[0042] Compared with the prior art, the traditional genetic algorithm uses binary coding and cannot express continuous variables such as inventory allocation proportion, while the present scheme directly maps the decision variable space through real number coding. The prior art usually uses a fixed weight fitness function, which cannot dynamically adjust the target weight according to the business priority, while the present scheme realizes dynamic weight configuration through the analytic hierarchy process. The conventional genetic algorithm uses fixed crossover and mutation probabilities, which can easily lead to premature convergence, while the present method maintains population diversity in the early evolution stage and enhances local search accuracy in the later stage through an adaptive mechanism.

[0043] Through the above technical means, the present application solves the problem of limited solution space caused by inaccurate variable expression in the multi-objective optimization process, and realizes the collaborative optimization of scheduling path, inventory distribution and pricing strategy. Through the dynamic weight distribution mechanism, the contradiction between cost control and service quality is balanced, and the business imbalance caused by single target optimization is avoided. The adaptive genetic operation effectively prevents the algorithm from falling into a local optimal solution, and quickly converges to a global optimal solution set under complex constraint conditions.

[0044] Further, a dynamic lease pricing adjustment step is proposed, a pricing factor system is constructed based on customer credit rating, lease quantity, lease period and historical cooperation frequency, customer credit rating is comprehensively evaluated through account performance record, settlement timeliness and tray loss rate; a gradient boosting regression model is used to train the pricing factors to generate a personalized pricing plan, and when the customer lease quantity exceeds the preset threshold or the lease period covers the industry off-season, the discount coefficient adjustment is automatically triggered.

[0045] Among them, the pricing factor system refers to taking customer credit rating, lease quantity, lease period and historical cooperation frequency as a variable set that affects the lease price, and specifically, data standardization methods can be used to convert each factor into a unified dimension value, and a pricing base parameter is formed by weighted combination, which is used to solve the problem that a single factor cannot reflect the comprehensive value of customers in traditional pricing.

[0046] Among them, the customer credit rating refers to the credit score calculated based on account performance record, settlement timeliness and tray loss rate, and specifically, the analytic hierarchy process can be used to determine the weight of each index, for example: performance record accounts for 40%, settlement timeliness accounts for 30%, and loss rate accounts for 30%, and a credit score of 0-100 is generated by weighted summation, which is used to distinguish the credit risk level of different customers. Among them, the gradient boosting regression model refers to an integrated learning algorithm, and specifically, the pricing factors and transaction prices in historical lease data can be used as a training set to iteratively optimize the residual to fit the nonlinear relationship of lease price, which is used to capture the complex relationship between quantity, period and price.

[0047] Among them, the preset threshold refers to the critical lease quantity that triggers the quantity discount, and specifically, it can be set to 1.5-2 times the average lease quantity in the industry according to historical order distribution, for example, when the single lease quantity exceeds 200 trays, the discount coefficient is activated, which is used to encourage large customers to increase the lease scale.

[0048] Among them, the industry off-season refers to the low demand period divided according to historical demand fluctuations, and specifically, time series analysis can be used to identify March-May as the logistics off-season every year, which is used to increase tray turnover rate through discounts in the demand trough period.

[0049] Specifically, the customer credit rating is quantitatively scored by the order completion rate in the account performance record, the median of the overdue days in the settlement timeliness, and the historical average of the pallet loss rate, forming a differentiated credit rating label. The lease quantity, cycle, and historical cooperation frequency are converted into 0-1 interval values through normalization processing, and are input into the gradient boosting regression model together with the credit rating. The model fits the non-linear relationship between the historical transaction price and each factor to generate a personalized offer including the base price and the floating interval. When the customer's lease quantity exceeds a preset threshold, the system automatically calls the discount coefficient calculation module, for example, a 5% discount is triggered when the lease quantity reaches 300; if the lease cycle covers the industry off-season, a discount coefficient for the off-season is added, for example, a 3% discount is additionally obtained for a 6-month service in March. Thus, the dynamic matching of price strategy, customer behavior, and market cycle is realized.

[0050] Compared with the prior art, the existing shared pallet lease pricing usually adopts fixed rate or simple segmented pricing, for example, a daily rental fee is uniformly charged according to the type of pallet, without considering the credit difference of customers and the characteristics of lease behavior. The present scheme builds a multi-dimensional pricing factor system, and the credit score, lease size and other variables are included in the model training, so that high-credit customers can obtain a lower base rate, long-term cooperation customers enjoy cycle cumulative discounts, and customers with excessive loss rate are subject to premium coefficients. At the same time, compared with the linear regression method, the gradient boosting regression model can more accurately reflect the non-linear interaction between price and multiple factors, for example, the price elasticity significantly changes in the 200-300 interval of the lease batch.

[0051] Through the above technical means, the present application solves the problem of revenue loss caused by rough customer classification in the traditional pricing mechanism. Through the quantitative evaluation of credit rating and lease behavior, high-risk customers bear higher costs, and high-quality customers obtain price incentives, thereby balancing platform revenue and customer satisfaction. The automatic discount mechanism for the industry off-season can effectively improve the utilization rate of idle pallets and avoid resource waste. In addition, the application of the gradient boosting regression model enables the pricing strategy to be continuously optimized based on historical data, adapting to market supply and demand changes.

[0052] Further, an order intelligent matching step is proposed, based on the pallet type, quantity, delivery address and use limit demand in the customer order, combined with the optimized inventory distribution data, the optimal outbound warehouse / point and carrier are automatically matched; in the matching process, the warehouse / point with the closest distance to the customer and the pallet availability rate ≥ 90% is preferentially selected, and the historical on-time delivery rate and transportation cost offer of the carrier are also referred to.

[0053] The tray availability rate ≥ 90% refers to the proportion of the number of available trays in the warehouse / point to the total inventory, which can be calculated by collecting the tray in-place state data through the real-time inventory monitoring system, and is used to ensure that the actual inventory of the warehouse / point meets the order demand. The historical on-time delivery rate refers to the proportion of orders completed on time by the carrier within a predetermined time period, which can be calculated by extracting the difference between the signing time and the promised time of historical orders of the carrier through the logistics data platform, and is used to evaluate the reliability of the transportation time of the carrier. The transportation cost quote refers to the fee standard provided by the carrier for a specific transportation distance and cargo volume, which can be obtained through the logistics bidding platform or the carrier pricing system, and is used to quantify the economy of the transportation service.

[0054] Specifically, after the customer submits an order containing the tray type, quantity, delivery address and usage period, the system analyzes the latitude and longitude coordinates of the delivery address based on the geographic information system, calculates the distance between the customer's location and each warehouse / point, and preferentially selects the three closest warehouses / points. The current inventory data of the warehouse / point is called to verify whether the tray availability rate reaches 90%. If the availability rate meets the standard, the warehouse / point is included in the candidate list. Then, the system retrieves the carrier information associated with the candidate warehouse / point from the carrier database, extracts its historical on-time delivery rate and current transportation cost quote. Through a weighted scoring model, the timeliness and economy of the candidate carrier are comprehensively evaluated, the optimal warehouse / point and carrier matching combination is generated, and the matching result is pushed to the dispatching system for execution.

[0055] Compared with the prior art, the traditional order matching relies on manual experience to select warehouses and carriers, and does not consider real-time inventory status and carrier service quality, which may lead to a deviation of the matching result from the optimal solution. The present method quantitatively evaluates the tray availability rate, transportation timeliness and cost parameters, constructs a multi-dimensional matching model, realizes dynamic and accurate adaptation of order demand and resource supply, and effectively eliminates subjective errors of manual decision-making.

[0056] Through the above technical solution, the present application solves the problem of order response delay caused by low efficiency of manual matching in the process of shared tray rental, optimizes the transportation path and carrier selection, and avoids additional scheduling costs caused by insufficient inventory or unstable carrier service. Through the automatic matching mechanism, the customer can obtain tray resources that meet the demand in the shortest time, and the overall logistics operating cost is reduced.

[0057] Further, an optimal recycling route is planned based on the recycling application data initiated by the member end, the geographic distribution of the points and the tray state data, the adjacent recycling demands are combined using a clustering algorithm to reduce the empty mileage of the recycling vehicles, and the recycling trays are detected by grade and distributed to the reuse, repair or scrap channels according to the detection results. The detection indicators include appearance integrity, structural strength and RFID signal stability.

[0058] wherein the clustering algorithm refers to a method of grouping recycling demand points with similar spatial positions, which can be specifically implemented by K-means algorithm or DBSCAN algorithm, a task set is formed by calculating the geographical distance between each recycling point, thereby reducing the repeated path of the vehicle. The appearance integrity detection refers to quantitatively evaluating the physical damage such as scratches and deformation on the surface of the tray through image recognition technology, which can be specifically implemented by a convolutional neural network model, and is used to judge whether the appearance of the tray meets the reuse standard. The structural strength detection refers to measuring the bearing capacity attenuation degree of the tray through a pressure sensor, which can be specifically implemented by a dynamic load testing device, and is used to evaluate whether the mechanical performance of the tray meets the safety threshold. The RFID signal stability detection refers to verifying the response frequency and data integrity of the electronic tag through a reader, which can be specifically implemented by a signal strength threshold judgment method, and is used to ensure the traceability of the tray in subsequent circulation.

[0059] Specifically, the recycling application data contains the position, quantity and state information of the to-be-recycled tray, and an initial recycling task set is generated in combination with the network point distribution data. The clustering algorithm divides adjacent recycling points into the same task group according to the geographical coordinates, forming an optimized multi-path recycling scheme. The recycling vehicle performs recycling operations according to the merged task group, avoiding empty running caused by round trips between scattered points. The tray completed recycling enters the detection link, the appearance integrity detection identifies the surface damage degree, the structural strength detection evaluates the bearing performance attenuation, and the RFID signal stability detection judges the readability of the chip. When all three indicators meet the standards, it is allocated to the reuse channel, only the structural strength does not meet the standards, it is allocated to the repair channel, and any one indicator seriously exceeds the standard, it is allocated to the scrap channel.

[0060] Compared with the prior art, the traditional recycling route planning relies on manual experience to divide regions, which is prone to problems such as path intersection or incomplete coverage, while the clustering algorithm automatically generates task grouping through spatial data analysis, improving the efficiency of route planning. The detection of the recycling tray in the prior art usually only performs visual inspection, lacking quantitative evaluation standards, while the present scheme establishes a hierarchical evaluation system through multi-dimensional detection indicators, improving the accuracy of classification processing.

[0061] Through the above technical means, the present application can effectively reduce the empty mileage of the recycling vehicle, reduce fuel consumption and carbon emissions, and at the same time improve the efficiency of tray classification and processing through an automated detection process, avoiding resource waste caused by manual misjudgment. The hierarchical detection mechanism ensures that trays in different states enter the appropriate processing channel, prolongs the service life of reusable trays, reduces maintenance costs, and ensures the timely exit of scrap trays.

[0062] Further, it is proposed to plan the optimal recovery route based on the member-end initiated recycling application data, the geographic distribution of network points, and the tray state data, to use clustering algorithms to combine adjacent recycling needs to reduce the empty mileage of recycling vehicles, and to conduct hierarchical detection of recycling trays and distribute them to the reuse, repair, or scrap channels according to the detection results. The detection indicators include appearance integrity, structural strength, and RFID signal stability.

[0063] Among them, the clustering algorithm refers to a method of grouping recycling demand points with similar spatial positions, which can be implemented by K-means algorithm or DBSCAN algorithm. By calculating the geographic distance between each recycling point, a task set is formed to reduce vehicle repeated paths. Appearance integrity detection refers to quantitative evaluation of physical damage such as scratches and deformation on the tray surface through image recognition technology. Specifically, it can be implemented by using a convolutional neural network model to determine whether the tray appearance meets the reuse standard. Structural strength detection refers to measuring the degree of attenuation of the tray's bearing capacity through a pressure sensor. Specifically, it can be implemented by using a dynamic load testing device to evaluate whether the tray's mechanical performance meets the safety threshold. RFID signal stability detection refers to verifying the response frequency and data integrity of the electronic tag through a reader. Specifically, it can be implemented by using a signal strength threshold judgment method to ensure the traceability of the tray in subsequent circulation.

[0064] Specifically, the recycling application data contains the location, quantity, and state information of the recycling trays, and the initial recycling task set is generated in combination with the network distribution data. The clustering algorithm divides adjacent recycling points into the same task group according to geographic coordinates to form an optimized multi-path recovery scheme. The recycling vehicles perform recovery operations according to the merged task groups to avoid empty driving caused by repeated trips between scattered points. The recycled trays enter the detection link, the appearance integrity detection identifies the degree of surface damage, the structural strength detection evaluates the bearing performance attenuation, and the RFID signal stability detection judges the readability of the chip. When all three indicators meet the standards, they are allocated to the reuse channel, only when the structural strength does not meet the standards, they are allocated to the repair channel, and when any indicator is seriously out of standard, they are allocated to the scrap channel.

[0065] Compared with the prior art, the traditional recovery route planning relies on manual experience to divide regions, which is prone to path intersection or incomplete coverage. The clustering algorithm automatically generates task grouping through spatial data analysis, improving the efficiency of route planning. The existing technology usually only conducts visual inspection of the recycling trays, lacking quantitative evaluation standards. The present scheme establishes a hierarchical evaluation system through multi-dimensional detection indicators to improve the accuracy of classification processing.

[0066] By the above technical means, the present application can effectively reduce the empty mileage of the recovery vehicle, reduce fuel consumption and carbon emissions, and at the same time improve the efficiency of tray classification processing through an automatic detection process, avoiding resource waste caused by manual misjudgment. The hierarchical detection mechanism ensures that trays in different states enter the appropriate processing channel, prolongs the service life of reusable trays, reduces maintenance costs, and ensures the timely withdrawal of scrap trays.

[0067] Further, an optimization effect evaluation step is proposed, an evaluation index system is established, and the index system includes cost reduction rate, inventory turnover improvement rate, order response time shortening rate, customer complaint rate and tray recycling rate; the actual operation data after optimization is compared and analyzed with the benchmark data before optimization, and the parameters and optimization target weights of the genetic algorithm are dynamically adjusted according to the evaluation results, so as to realize continuous iterative optimization of the leasing scheme.

[0068] Among them, the evaluation index system refers to a multi-dimensional index set for quantitatively evaluating the optimization effect, and specifically, the cost reduction rate can be used to reflect the change of transportation cost, the inventory turnover improvement rate can be used to measure inventory efficiency, the order response time shortening rate can be used to evaluate service speed, the customer complaint rate can be used to reflect customer satisfaction, and the tray recycling rate can be used to represent resource utilization efficiency. These indicators are collected and calculated by the data statistics module to provide a basis for subsequent parameter adjustment.

[0069] Among them, dynamically adjusting the parameters and optimization target weights of the genetic algorithm means that the crossover probability, mutation probability and weight coefficients of each optimization target of the genetic algorithm are modified according to the evaluation results, and specifically, the parameters can be automatically adjusted according to the preset rules by using an adaptive algorithm, for example: when the inventory turnover improvement rate does not meet the expectation, the weight coefficient of the inventory allocation strategy is increased. This adjustment mechanism can avoid the local optimal problem caused by fixed parameters of the algorithm.

[0070] Specifically, the optimization effect evaluation step forms a closed-loop feedback mechanism through the construction of multi-dimensional indexes. The evaluation index system converts the operation data of different dimensions into quantifiable evaluation standards, for example, by comparing the order response time before and after optimization, it can be judged whether the dispatch path optimization is effective. The difference between the actual operation data and the benchmark data is analyzed regularly, for example: when the customer complaint rate rises, it indicates that the current pricing strategy or service process has deficiencies. Based on the difference analysis results, the parameters of the genetic algorithm are adjusted, the transportation cost weight is reduced to prioritize customer satisfaction, or the mutation probability is increased to enhance the search ability of the algorithm. This dynamic adjustment enables the optimization process to respond to business changes in real time, for example, increasing the optimization priority of inventory turnover rate in the sales peak season, and focusing on the improvement of tray recycling rate during the maintenance peak period.

[0071] In the specific embodiment, the benchmark data of the evaluation index system can be set as the historical average value of the first three months before optimization, and the actual operation data is collected in real time through the Internet of Things equipment. The comparative analysis can use difference rate calculation, for example, cost reduction rate = (benchmark transportation cost- actual transportation cost) / benchmark transportation cost x 100%. The parameter adjustment strategy can be set as: when the order response time shortening rate is less than 5%, the population size of the genetic algorithm is expanded by 20%; when the tray recycling rate decreases for two consecutive periods, the priority weight of the maintenance channel is increased.

[0072] Compared with the prior art, the existing shared tray rental scheme usually uses fixed parameters after single optimization, and lacks continuous tracking and feedback adjustment of actual effect. For example: after the traditional system completes path optimization, it does not reevaluate the change of transportation cost according to subsequent order fulfillment data, resulting in the decay of optimization effect over time. While the present scheme can identify optimization deviation caused by supply chain fluctuations and inventory allocation imbalance caused by holiday order surge through the establishment of a regular evaluation mechanism, and then trigger targeted adjustment of algorithm parameters to form a positive cycle of continuous optimization.

[0073] Through the above technical means, the present application solves the problem of insufficient adaptability of the scheme caused by the fixation of optimization parameters in the prior art. By quantitatively evaluating the actual operation effect and dynamically correcting the algorithm parameters, it ensures that the optimization strategy always matches the real-time business status. For example: when a new warehouse node is added in a certain area, the evaluation mechanism can quickly identify the inventory turnover rate data of the node and adjust the inventory allocation constraint conditions of the genetic algorithm accordingly. This closed-loop feedback mechanism effectively maintains the continuity of the optimization effect and avoids the failure of the scheme caused by environmental changes.

[0074] As shown in Figure 2 The present embodiment also proposes a shared tray rental scheme optimization system, which includes a data acquisition and preprocessing module 10, a digital twin modeling module 20, a multi-objective optimization module 30, and a function execution module 40.

[0075] As shown in Figure 3 The function execution module 40 includes a rental pricing dynamic adjustment unit 401, an order intelligent matching unit 402, a recycling optimization unit 403, an abnormal order intelligent processing unit 404, and an optimization effect evaluation unit 405.

[0076] The data acquisition and preprocessing module 10 is an edge computing node deployed in the warehouse or network point, which collects tray in-out warehouse data, device state data and environment data through sensors and 5G communication modules. Specifically, it can use an industrial-grade Internet of Things gateway to realize data deduplication and outlier filtering, solving the problem of low efficiency of traditional manual collection.

[0077] The digital twin modeling module 20 refers to binding the RFID tag with the virtual model, predicting the life of the pallet through time series analysis, and specifically can use Apache Kafka to realize real-time data stream synchronization, solving the problem of disconnection between physical assets and digital information.

[0078] The multi-objective optimization module 30 refers to using a genetic algorithm with real number coding, and the adaptive crossover probability is dynamically adjusted through the number of evolution generations. Specifically, it can use the NSGA-II framework to balance transportation cost and inventory turnover, solving the local optimal problem caused by fixed parameters.

[0079] The lease pricing dynamic adjustment unit 401 refers to training pricing factors based on a gradient boosting regression model. Specifically, it can use the XGBoost algorithm to process customer credit rating and lease period variables, solving the problem of single traditional pricing strategy.

[0080] The order intelligent matching unit 402 refers to screening the optimal network point by combining geographic information system and inventory availability. Specifically, it can use the Dijkstra algorithm to calculate the distribution path, solving the problem of response delay of manual matching.

[0081] Specifically, the data collection and preprocessing module 10 inputs the cleaned inventory data into the digital twin modeling module 20 to generate a virtual model containing the degree of pallet wear and position coordinates.

[0082] The multi-objective optimization module 30 receives the pallet state data output by the model and generates scheduling paths and pricing strategies through improved genetic algorithms. The order intelligent matching unit 401 in the function execution module 40 selects the warehouse closest to the customer and with the required availability rate as the priority based on the optimization results, and simultaneously calls the recycling optimization unit 403 to perform clustering analysis on the returned pallets to plan the recycling route. The abnormal order intelligent processing unit 404 monitors the order status in real time through a random forest model and automatically triggers a financial reminder process when settlement abnormalities are detected. The optimization effect evaluation unit 405 periodically compares indicators such as inventory turnover rate and feeds back to the multi-objective optimization module to realize parameter iteration.

[0083] Compared with the prior art, the traditional system relies on manual experience to adjust the leasing strategy and cannot respond to inventory changes in real time. The prior art uses fixed genetic algorithm parameters, resulting in unstable optimization results, while the present scheme dynamically balances search efficiency through adaptive crossover probability. The existing recycling process does not consider geographical location clustering, while the present scheme uses the K-means algorithm to merge adjacent recycling points, reducing empty running rate. The existing pricing model does not integrate customer credit data, while the present scheme realizes differentiated pricing through a machine learning model.

[0084] Through the above technical means, the dynamic optimization of the tray scheduling path is realized, and the empty load rate of the transport vehicle is reduced. Through the digital twin model, the physical tray state is synchronized, and the inventory mismatch caused by information delay is avoided. The machine learning driven anomaly detection mechanism shortens the order exception response time. The intelligent matching strategy based on geographic location and inventory availability improves the order processing efficiency. The closed-loop optimization evaluation mechanism ensures the continuous optimization of system parameters, solving the problem of insufficient iteration capability of traditional systems.

[0085] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for optimizing a shared pallet rental scheme, characterized in that, Includes the following steps: Acquire historical leasing data, real-time inventory data, and forecasted demand data for each node in the supply chain; Based on RFID technology, the entire lifecycle trajectory data of pallets is collected to establish a digital twin model of pallet circulation, which maps the entire process status of pallets from warehousing, leasing, transportation, recycling to repair / scrapping; The improved genetic algorithm is used to perform multi-objective optimization of pallet scheduling paths, inventory distribution, and rental pricing.

2. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, In the step of acquiring data from each node of the supply chain, edge computing nodes are deployed at warehouses / outlets to collect pallet entry and exit data, equipment operating status data, and environmental temperature and humidity data in real time. The pre-processed data is uploaded to the cloud platform via a 5G communication module. The pre-processing process includes data deduplication, outlier removal, and format standardization. Outlier removal is based on the 3σ principle to identify loss rates and inventory quantity data that exceed the normal range.

3. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, In the step of establishing a digital twin model of pallet circulation, the unique pallet identifier collected by RFID is bound to the virtual pallet in the digital twin model, and the physical status parameters of the pallet are synchronized in real time. The physical state parameters include wear level, load history, location coordinates and associated order information. The remaining service life of the pallet is predicted through time series analysis. When the remaining service life is lower than the first preset threshold, a maintenance reminder or scrap warning is automatically triggered.

4. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, In the improved genetic algorithm optimization step, the encoding method adopts real number encoding, and the pallet scheduling path, inventory allocation ratio and pricing coefficient are used as gene segments; The fitness function is constructed as follows: transportation cost weight coefficient × transportation cost value + inventory turnover rate weight coefficient × (1 - inventory turnover cycle) + customer satisfaction weight coefficient × customer satisfaction value. The weight coefficients are determined by the analytic hierarchy process. In genetic operations, adaptive crossover and mutation probabilities are introduced and dynamically adjusted according to the number of generations of population evolution to avoid the algorithm getting trapped in local optima.

5. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, This includes dynamic adjustment steps for lease pricing: a pricing factor system is built based on customer credit rating, lease volume, lease period, and historical cooperation frequency. Customer credit rating is comprehensively assessed through account performance records, settlement timeliness, and pallet loss rate. A gradient boosting regression model is used to train the pricing factors and generate personalized pricing schemes. When the customer's rental volume exceeds the second preset threshold or the rental period covers the industry's off-season, the discount coefficient is automatically adjusted.

6. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, This includes an intelligent order matching process: based on the pallet type, quantity, delivery address, and usage period requirements in customer orders, combined with optimized inventory distribution data, it automatically matches the optimal outbound warehouse / outlet and carrier; During the matching process, priority is given to warehouses / outlets that are closest to the customer and have pre-set pallet availability, while also taking into account the carrier's historical on-time delivery rate and transportation cost quotes.

7. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, The recycling optimization steps include: planning the optimal recycling route based on recycling application data initiated by members, the geographical distribution of recycling points, and pallet status data; and using clustering algorithms to merge adjacent recycling requests to reduce the empty mileage of recycling vehicles. The recycling pallets are graded and tested, and then allocated to the reuse, repair or disposal channels according to the test results. The test indicators include appearance integrity, structural strength and RFID signal stability.

8. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, This includes intelligent processing steps for abnormal orders: real-time monitoring of rental order status through machine learning models to identify abnormal situations such as order delays, pallet damage, and settlement anomalies; The machine learning model is generated using the random forest algorithm, trained based on historical abnormal order data. Corresponding handling strategies are implemented for different types of abnormal triggers. When an order is delayed, the subsequent scheduling plan is automatically adjusted. When a pallet is damaged, the claims process is initiated and a replacement pallet is provided. When a settlement is abnormal, a reminder message is pushed to the finance module.

9. The method for optimizing a shared pallet rental scheme according to claim 1, characterized in that, This includes the optimization effect evaluation steps: establishing an evaluation indicator system, which includes cost reduction rate, inventory turnover improvement rate, order response time reduction rate, customer complaint rate, and pallet recycling rate; Regularly compare and analyze the optimized actual operating data with the baseline data before optimization, and dynamically adjust the parameters of the genetic algorithm and the optimization target weights based on the evaluation results to achieve continuous iterative optimization of the leasing scheme.

10. A shared pallet rental scheme optimization system, used to execute a shared pallet rental scheme optimization method according to any one of claims 1-9, characterized in that, include: The data acquisition and preprocessing module is used to acquire historical leasing data, real-time inventory data, and forecasted demand data from each node of the supply chain. The digital twin modeling module, based on the full lifecycle trajectory data of the pallet collected by RFID technology, establishes a digital twin model of pallet circulation, binds the unique pallet identifier collected by RFID to the virtual pallet in the digital twin model, and synchronizes the physical status parameters of the pallet in real time and maps the status of the entire process. The multi-objective optimization module uses an improved genetic algorithm to perform multi-objective optimization of pallet scheduling paths, inventory distribution, and rental pricing; The system also includes a function execution module, comprising a rental pricing dynamic adjustment unit, an order intelligent matching unit, a recycling optimization unit, an abnormal order intelligent processing unit, and an optimization effect evaluation unit; the function execution module is configured to execute the steps of the shared pallet rental scheme optimization method according to any one of claims 5-9.

Images