A method for inventory destocking and risk-controllable cross-regional scheduling of durable goods

Abstract

The application discloses a kind of inventory backlog of durable goods and risk controllable cross-regional scheduling method, belong to data scheduling technical field, comprising the following steps: S1, the data set of whole area is collected, and pretreatment is carried out;S2, using first condition or second condition determines substantial backlog inventory, and generates corresponding scheduling priority;S3, using product value attribute comprehensive score, transportability comprehensive score and the total risk score corresponding to transport loss consequence comprehensive score determine product risk grade;S4, using first priority, second priority and third priority carry out optimal scheduling execution.The application realizes core index real-time monitoring, abnormal automatic early warning, performance appraisal data automatic statistics, rule dynamic optimization, complete whole-process closed-loop management.

Description

Technical Field

[0001] This invention belongs to the field of data scheduling technology, specifically relating to a method for resolving inventory backlog of durable goods and cross-regional scheduling with controllable risks. Background Technology

[0002] Durable goods such as machinery and home appliances are characterized by high unit price, large size and weight, high cross-regional transportation costs, easy damage from bumps and collisions during long-distance transportation, and large capital tied up in inventory. The core pain points of their supply chain inventory management are concentrated in four aspects: uneven inventory turnover, serious local backlog, lack of unified standards for cross-regional dispatch, and difficulty in controlling transportation losses of high-risk products.

[0003] Currently, inventory management in the industry generally follows the core logic of "order fulfillment and stockout replenishment," and cross-regional transfers are only triggered when there is an inventory gap in downstream warehouses. The elimination of backlogged inventory is not taken as the core objective of the management, resulting in a large amount of backlogged inventory tying up the company's working capital for a long time. At the same time, the additional losses caused by the indiscriminate cross-regional transfer of high-risk products further increase the overall cost of the supply chain.

[0004] Existing technical solutions struggle to balance the dual objectives of "rapidly reducing backlogged inventory" and "controlling overall scheduling costs." They also lack a standardized mechanism for determining backlogged inventory and scheduling control rules adapted to product risk levels. Consequently, they fail to address the root causes of inventory backlog and scheduling losses in the machinery / home appliance durable goods industry. There is an urgent need for an automated scheduling technology solution that is adapted to industry characteristics, can be standardized and implemented, and balances backlog reduction, cost control, and risk management.

[0005] The current common operating model for small and medium-sized machinery / home appliance companies is a combination of manual experience-based independent regional inventory management and passive cross-regional transfer solutions. Under this model, each regional / area warehouse independently manages its own inventory, with warehouse managers judging backlogs based on manual experience. Cross-regional transfer requests are only initiated to other regional warehouses when there are order shortages or inventory gaps in the warehouse within the same region. The transfer is executed after manual confirmation between the two parties, without a unified overall control and scheduling system or standardized rules.

[0006] However, the drawbacks are as follows: 1. There are no standardized rules for determining backlog, relying entirely on subjective human judgment. Inconsistent judgment standards across different regions mean that large amounts of backlog cannot be identified and disposed of in a timely manner, resulting in long-term tied-up working capital and low inventory turnover efficiency. 2. The dispatching model is passive replenishment, completely neglecting backlog resolution as a dispatching objective. It cannot achieve cross-regional digestion of backlog through coordinated regional dispatching, failing to address the backlog problem at its root. 3. The risk of product transportation losses is not considered. High-risk products are dispatched across regions indiscriminately, leading to high product loss rates due to long-distance transportation, resulting in additional cost losses and fulfillment risks. 4. There are no cost control rules for cross-regional transfers. Manual communication and decision-making can easily lead to ineffective transportation and unreasonable routes, resulting in high overall supply chain costs and extremely low management efficiency. 5. Backlog determination relies solely on human experience and monthly inventory data, failing to consider short-term data fluctuations caused by peak and off-peak seasons and major promotional activities in the machinery / home appliance durable goods industry. This easily leads to misjudgments or omissions of backlog, making it impossible to accurately identify substantial backlog inventory that truly poses financial risks.

[0007] Automated inventory scheduling systems based on order fulfillment are commonly used solutions for medium and large enterprises. Related technologies can be found in patent literature related to warehouse management systems (WMS) and supply chain scheduling systems. This solution integrates inventory and order data across warehouses, automatically triggering cross-regional transfers based on downstream order demand and inventory gaps, matching the transferring warehouse with the transportation route. The core objective is to ensure timely order fulfillment and on-the-shelf availability. Some solutions incorporate basic optimization logic for transportation costs. However, its drawbacks are as follows: 1. The scheduling logic focuses solely on order fulfillment and stockout replenishment, failing to prioritize inventory clearance and thus hindering proactive and coordinated clearance of inventory across the entire region. It can only address stockouts, not inventory backlogs. 2. It lacks a standardized mechanism for identifying and classifying inventory backlogs tailored to the characteristics of durable goods like machinery and home appliances. This prevents automated identification and prioritization of backlogs and hinders automated scheduling for inventory clearance. 3. It fails to consider the differences in transportation losses for products with varying risk levels and lacks differentiated scheduling authority control rules. The risk of losses from cross-regional scheduling of high-risk products cannot be mitigated at the source, leading to persistently high product loss costs. 4. It cannot simultaneously achieve the dual objectives of "prioritizing inventory clearance" and "lowest overall cost," only enabling single-dimensional cost optimization and failing to balance the core needs of inventory clearance and cost control. 5. Using only the single indicator of monthly turnover rate to determine backlog, without establishing a multi-dimensional verification mechanism, cannot eliminate abnormal fluctuations in monthly data caused by industry peak and off-peak seasons and promotional cycles. The accuracy of backlog determination is insufficient, and misjudgments are prone to occur, making it impossible to support precise backlog resolution and scheduling. Summary of the Invention

[0008] To address the above problems, this invention proposes a method for resolving inventory backlogs and managing risks in cross-regional scheduling of durable goods.

[0009] The technical solution of this invention is: a method for resolving inventory backlog and controlling the risk of cross-regional dispatching of durable goods, comprising the following steps:

[0010] S1. Collect the dataset from the entire region and perform preprocessing;

[0011] S2. Based on the preprocessed dataset, determine the substantial backlog of inventory using the first or second condition, and generate the corresponding scheduling priority; the first condition is set by the monthly inventory turnover rate; the second condition is set by the inventory percentage.

[0012] S3. Based on the scheduling priority corresponding to the substantial backlog of inventory, determine the product risk level using the total risk score corresponding to the comprehensive score of product value attributes, comprehensive score of transportation vulnerability, and comprehensive score of transportation loss consequences.

[0013] S4. Based on the product risk level, perform optimal scheduling execution using the first priority, second priority, and third priority; the second priority is obtained by solving the objective function to obtain the scheduling route.

[0014] Furthermore, in S1, the dataset includes inventory data, order data, basic product data, and transportation route data.

[0015] Furthermore, S2 includes the following sub-steps:

[0016] S21. Perform preliminary screening of the preprocessed dataset using the basic judgment values;

[0017] S22. Based on the preliminary screening of the dataset, determine the actual backlog of inventory;

[0018] S23. Divide the actual backlog of inventory and determine the corresponding scheduling priority.

[0019] Furthermore, in S22, the dataset that meets the first condition or the second condition is treated as substantial backlog inventory;

[0020] The first condition is that the monthly inventory turnover rate is lower than the preset backlog threshold for the corresponding product category for two consecutive calendar months.

[0021] The second condition is that the inventory percentage of the preset age segment exceeds the preset percentage threshold.

[0022] Furthermore, in S23, the scheduling priority includes first-level backlog, second-level backlog, and third-level backlog;

[0023] Level 1 backlog is defined as follows: the turnover rate lags behind the judgment threshold by 50% or more, and the inventory capital ratio of a single product exceeds the first threshold of the total inventory capital ratio in the region.

[0024] Level 2 backlog is defined as follows: the turnover rate lags behind the judgment threshold by 30%-50%, and the inventory capital ratio of a single product exceeds the second threshold of the total inventory capital ratio in the region.

[0025] Level 3 backlog is defined as a turnover rate that lags behind the judgment threshold by no more than 30%.

[0026] Furthermore, in S3, the product risk level is determined as follows: the comprehensive score of product value attributes, the comprehensive score of transportation vulnerability, and the comprehensive score of transportation loss consequences are weighted and integrated to obtain the total risk score; if the total risk score is less than 3, the product risk level is low risk; if the total risk score is greater than or equal to 3 and less than 5, the product risk level is medium risk; if the total risk score is greater than or equal to 5, the product risk level is high risk.

[0027] Overall Risk Score The expression is:

[0028] ;

[0029] ;

[0030] ;

[0031] ;

[0032] in, The weighting of the overall score for product value attributes. The weighting of the overall score for transport fragility. The weighting of the overall score for the consequences of transportation losses. A comprehensive score for the product's value attributes. For the comprehensive score of transport vulnerability, A comprehensive score is given for the consequences of transportation losses. This refers to the unit purchase price of a single product. Historical transport damage rate For product packaging protection level, For the product's IP protection rating, This represents the percentage of repair costs per unit damaged. For long-distance transportation defect rate.

[0033] Furthermore, in S4, optimal scheduling execution includes first priority, second priority, and third priority.

[0034] Furthermore, the first priority is specifically: designate the region to which the product belongs as the scheduling region.

[0035] Furthermore, the second optimization stage specifically involves: determining edge weights based on the undirected graph model, determining the objective function based on the edge weights, and solving the problem using the improved Dijkstra path optimization method to obtain the scheduling route;

[0036] edge weight The expression is:

[0037] ;

[0038] in, The weighting coefficient for the standardized unit transportation cost. The weighting coefficients for the standardized cost of breakage risk. The weighting coefficient for standardized transportation time costs. The standardized unit transportation cost, For the standardized cost of breakage risk, For standardized transportation time costs, For the undirected graph model, the edges correspond to the transportation routes between warehouses i and j;

[0039] The expression for the objective function is: ,in, This represents the total weight of the path. For path.

[0040] Furthermore, the third priority is to generate a clearing list and send it to the operations department for processing.

[0041] The beneficial effects of this invention are:

[0042] (1) This invention connects to the WMS system and order management system of all warehouses in the region to complete the collection, cleaning, format unification and standardized storage of multi-source data, and provides data support for the whole process.

[0043] (2) The present invention has a three-layer judgment model with built-in basic threshold screening, multi-dimensional review of substantial backlog and dual-dimensional classification of backlog level. It automatically completes the initial screening, accurate review, classification and priority sorting of backlog inventory, which can effectively avoid misjudgment of monthly data fluctuations caused by industry off-seasons and promotion cycles, and output an accurate list of substantial backlog inventory.

[0044] (3) The present invention has built-in differentiated management and control rules for three types of products, realizing automatic identification of risk level, automatic verification of scheduling authority and automatic interception of high-risk operations, and completing closed-loop risk management and control.

[0045] (4) The present invention incorporates a three-level priority scheduling logic and an undirected graph cost-optimal path matching algorithm to automatically complete intra-regional matching, cross-regional overall scheduling, instruction generation and approval control, and achieve the dual objectives of backlog resolution and cost optimization.

[0046] (5) This invention enables real-time monitoring of core indicators, automatic early warning of anomalies, automatic statistics of performance evaluation data, and dynamic optimization of rules, thus completing closed-loop management of the entire process. Attached Figure Description

[0047] Figure 1 A flowchart for a cross-regional scheduling method to reduce inventory backlog and manage risk for durable goods. Detailed Implementation

[0048] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0049] Inventory turnover rate: A core indicator for measuring inventory turnover efficiency. Monthly inventory turnover rate = monthly cost of goods sold ÷ average monthly inventory balance, where average monthly inventory balance = (beginning inventory balance + ending inventory balance) ÷ 2.

[0050] Substantial backlog: refers to inventory that, after excluding short-term factors such as peak and off-peak seasons and promotional activities that cause monthly turnover fluctuations, is determined through standardized review to have the risk of long-term turnover stagnation and capital tied up. This is the core target of the backlog disposal and scheduling in this plan.

[0051] Inventory age structure: refers to the proportion of inventory quantity / value in different age segments of a single SKU product inventory to the total inventory. It is a core indicator for measuring inventory health and judging the risk of backlog.

[0052] Overstocked inventory: refers to inventory that exceeds the normal turnover cycle, ties up a lot of capital, and has no stable demand for delivery in the long term. In this solution, it is determined by standardized quantitative thresholds.

[0053] Three categories of risky products: A tiered system based on transportation loss risk, value attributes, and fragility of durable goods such as machinery and home appliances. Regional tiered management: Centered on the central control center, operational areas are divided according to business regions to achieve a hierarchical inventory management model of "prioritizing inventory disposal within each area and providing overall support across areas."

[0054] Undirected graph scheduling: This technique abstracts all warehouses in the region as nodes in an undirected graph, and the transportation routes between warehouses as edges. It uses transportation cost and scheduling priority as weights to achieve a globally optimal match between backlogged inventory and demand gaps.

[0055] SKU: StockKeeping Unit, refers to the smallest salable unit used to uniquely identify, distinguish, manage, and track product inventory. Different SKUs correspond to different products if they differ in specifications, models, sizes, colors, configurations, or versions. SKUs are the basic identifying units for product inventory management, order processing, and product differentiation.

[0056] like Figure 1 As shown, this invention provides a method for resolving inventory backlogs and controlling the risks of cross-regional scheduling of durable goods, comprising the following steps:

[0057] S1. Collect the dataset from the entire region and perform preprocessing;

[0058] S2. Based on the preprocessed dataset, determine the substantial backlog of inventory using the first or second condition, and generate the corresponding scheduling priority; the first condition is set by the monthly inventory turnover rate; the second condition is set by the inventory percentage.

[0059] S3. Based on the scheduling priority corresponding to the substantial backlog of inventory, determine the product risk level using the total risk score corresponding to the comprehensive score of product value attributes, comprehensive score of transportation vulnerability, and comprehensive score of transportation loss consequences.

[0060] S4. Based on the product risk level, perform optimal scheduling execution using the first priority, second priority, and third priority; the second priority is obtained by solving the objective function to obtain the scheduling route.

[0061] The problem solved by this invention is:

[0062] 1. To address the technical problems in the existing technology of machinery / home appliance durable goods industry, such as the lack of standardized and quantifiable mechanisms for the identification and graded disposal of backlogged inventory, the reliance on manual labor for backlogged inventory identification and untimely disposal, resulting in low inventory turnover efficiency and long-term tying up of working capital.

[0063] 2. Addressing the technical shortcomings of existing inventory scheduling technologies, which focus solely on order fulfillment and fail to simultaneously achieve the dual objectives of prioritizing the reduction of backlogged inventory and minimizing overall scheduling costs, thus failing to fundamentally solve the problem of inventory backlog.

[0064] 3. Address the core technical pain point that existing scheduling technologies fail to consider the differences in transportation losses for products with different risk levels and lack differentiated scheduling and control rules, resulting in persistently high cross-regional scheduling loss rates for high-risk products.

[0065] 4. To address the problems of existing inventory management models where each region operates independently without a unified overall scheduling system, cross-regional goods transfers lack standardized processes, resulting in low management efficiency and uncontrollable overall costs, we provide a standardized, fully automated scheduling solution that is adapted to the characteristics of the machinery / home appliance durable goods industry.

[0066] 5. Address the technical shortcomings of existing technologies that rely solely on a single monthly turnover rate to determine backlogged inventory, failing to consider short-term data fluctuations caused by peak and off-peak seasons and promotional activities in the machinery / home appliance durable goods industry. This leads to misjudgments and omissions of backlogged inventory and makes it difficult to accurately identify substantial backlogged inventory. Improve the accuracy and scenario adaptability of backlog determination.

[0067] In this embodiment of the invention, in S1, the dataset includes inventory data, order data, basic product data, and transportation route data.

[0068] Data collection dimensions: beginning / end inventory quantity of a single product, beginning / end inventory balance of a single product, real-time inventory quantity of a single product, inbound time of a single product, inventory age segment data, historical monthly turnover rate data, monthly sales outbound quantity, monthly sales cost, product unit price, product risk level, warehouse region, inter-warehouse transportation cost and timeliness, order demand and inventory gap data.

[0069] Preprocessing: Standardize data format and statistical standards, remove outliers and duplicate data, complete missing data, and classify and store data by region, product category, and risk level to provide standardized data support for subsequent judgment and scheduling.

[0070] In this embodiment of the invention, S2 includes the following sub-steps:

[0071] S21. Perform preliminary screening of the preprocessed dataset using the basic judgment values;

[0072] S22. Based on the preliminary screening of the dataset, determine the actual backlog of inventory;

[0073] S23. Divide the actual backlog of inventory and determine the corresponding scheduling priority.

[0074] For ordinary value machinery / home appliances, a corresponding backlog judgment threshold is set based on the average monthly inventory turnover rate of the industry category.

[0075] For high-value machinery / home appliance products, based on their characteristics of high capital occupation and low turnover frequency, a more lenient threshold for determining overstock is set compared to ordinary products.

[0076] The system has a built-in dynamic threshold configuration module, which allows enterprises to customize and adjust the judgment threshold according to their own business characteristics, product turnover patterns, and business objectives, thus adapting to the differentiated management needs of different enterprises.

[0077] The system performs an initial screening of all SKU inventory and identifies a list of inventory items awaiting review whose monthly turnover rate is lower than the corresponding initial screening threshold.

[0078] In this embodiment of the invention, in S22, the dataset that satisfies the first condition or the second condition is regarded as substantial backlog inventory;

[0079] The first condition is that the monthly inventory turnover rate is lower than the preset backlog threshold for the corresponding product category for two consecutive calendar months.

[0080] The second condition is that the inventory percentage of the preset age segment exceeds the preset percentage threshold.

[0081] In this embodiment of the invention, based on the dual dimensions of "turnover lag and inventory capital ratio", the inventory determined to be backlogged is divided into 3 levels. The greater the turnover lag and the higher the inventory capital ratio, the higher the backlog level and the higher the scheduling priority.

[0082] In S23, scheduling priorities include Level 1 backlog, Level 2 backlog, and Level 3 backlog;

[0083] Level 1 backlog is defined as follows: the turnover rate lags behind the judgment threshold by 50% or more, and the inventory capital ratio of a single product exceeds the first threshold of the total inventory capital ratio in the region.

[0084] Level 2 backlog is defined as follows: the turnover rate lags behind the judgment threshold by 30%-50%, and the inventory capital ratio of a single product exceeds the second threshold of the total inventory capital ratio in the region.

[0085] Level 3 backlog is defined as a turnover rate that lags behind the judgment threshold by no more than 30%.

[0086] In this embodiment of the invention, in S3, determining the product risk level specifically involves: weighting and fusing the comprehensive score of product value attributes, the comprehensive score of transportation vulnerability, and the comprehensive score of transportation loss consequences to obtain a total risk score; if the total risk score is less than 3, the product risk level is low risk; if the total risk score is greater than or equal to 3 and less than 5, the product risk level is medium risk; if the total risk score is greater than or equal to 5, the product risk level is high risk.

[0087] The overall risk score The expression is:

[0088] ;

[0089] ;

[0090] ;

[0091] ;

[0092] in, The weighting of the overall score for product value attributes. The weighting of the overall score for transport fragility. The weighting of the overall score for the consequences of transportation losses. A comprehensive score for the product's value attributes. For the comprehensive score of transport vulnerability, A comprehensive score is given for the consequences of transportation losses. This refers to the unit purchase price of a single product. Historical transport damage rate For product packaging protection level, For the product's IP protection rating, This represents the percentage of repair costs per unit damaged. For long-distance transportation defect rate.

[0093] The final total risk score for a single SKU product ranges from 1 to 7. This ensures that the weighting coefficients are normalized and the calculation results are standardized and comparable; all scoring calculations are completed automatically by the system without human intervention, ensuring the objectivity and consistency of the judgment results.

[0094] Based on the industry characteristics of machinery / household appliances durable goods, three core dimensions and six quantifiable secondary indicators were selected, without subjective qualitative descriptions. The specific definitions are shown in Table 1.

[0095] Table 1

[0096] Product Value Attributes Purchase price per unit (including tax) The unit purchase cost of a single SKU product is based on data from the product's basic file in the enterprise's ERP system, directly reflecting the risk of capital tied up in the product. Transportation vulnerability Historical transport damage rate The breakdown of the number of damaged products for this SKU during cross-regional transportation over the past six months, divided by the total number of shipments, is based on return receipts and damage records from the TMS transportation management system. This data directly reflects the probability of damage during product transportation. Transportation vulnerability Product packaging protection level Packaging protection levels are classified according to national packaging and transportation standards into four levels: Level 1, Level 2, Level 3, and Level 4. The higher the level, the weaker the protective capability and the higher the vulnerability. Data is sourced from basic product files. Transportation vulnerability Product IP Protection Rating The product casing's dust and water protection rating (IP rating) is as follows: the lower the IP rating, the higher the probability of damage due to impacts or moisture during transportation. This data is sourced from the product's basic documentation. Consequences of transportation losses Percentage of repair costs per unit of damaged unit The higher the ratio of average repair cost per unit of this SKU divided by the purchase price per unit, the greater the economic loss caused by transportation damage. The data comes from repair work orders in the company's after-sales system. Consequences of transportation losses Long-distance transportation defect rate The data, calculated as follows (within the past 6 months), represents the number of customer returns / exchanges for this SKU after long-distance inter-provincial transportation, divided by the total number of shipments. This data, sourced from the order and after-sales system, reflects the overall loss risk associated with long-distance transportation.

[0097] This invention uses the Analytic Hierarchy Process (AHP) to determine the weight coefficients of each dimension. Considering the characteristics of the home appliance / machinery durable goods industry, transportation vulnerability has the greatest impact on the final risk, followed by the consequences of transportation loss, and finally the product value attributes. The weight coefficients are set as shown in Table 2, and enterprises can customize and adjust them.

[0098] Table 2

[0099] Dimension 2 (Transportation Vulnerability) Weighting coefficient: W2 = 0.40 Dimension 3 (Consequences of Transportation Losses) Weighting coefficient: W3 = 0.35

[0100] Based on the transportation loss risk, fragility, and value attributes of durable goods such as machinery and household appliances, a differentiated management and control system is established for three risk categories: A, B, and C. Risk management rules are embedded into the entire scheduling process upfront to mitigate transportation loss risks from the source. Specific rules are as follows:

[0101] Category A (Low-risk) products: Conventional mechanical parts and home appliances that are not easily damaged and have low transportation requirements. They are granted normal dispatch permissions throughout the region without any additional dispatch restrictions.

[0102] Category B (Medium Risk) Products: Products with certain transportation protection requirements and prone to minor bumps and damage. The scale of cross-regional single transfers is limited. The quantity of a single transfer cannot exceed the preset percentage threshold of the average monthly outbound volume of the product in the region. The threshold can be customized to avoid secondary backlogs and batch losses caused by sales fluctuations.

[0103] Category C (High-Risk) Products: Large machinery and precision home appliances that are high-value, easily damaged, and have an extremely high risk of loss during long-distance transportation. The system automatically locks their cross-regional transfer permissions, allowing them to be sold only within the local region, and prioritizing the triggering of the local rapid clearance mechanism.

[0104] This invention includes an implementation scenario involving a major electrical appliance distribution company. The company's business covers 6 provincial-level operating regions and 21 regional warehouses, primarily dealing in major appliances such as refrigerators, washing machines, and air conditioners, as well as related accessories. This algorithm automates the determination of product risk levels. The specific calculation process is as follows:

[0105] Case 1: Common household appliance parts (air conditioner filter).

[0106] Indicator data: Unit price 150 yuan (1 point), historical transportation damage rate 0.05% (1 point), packaging protection level 4 (1 point), IP protection level IP54 (1 point), maintenance cost as a percentage of total cost (1 point), long-distance transportation defect rate 0.1% (1 point).

[0107] Dimensional scoring: R1=1 point, R2=(1+1+1)÷3=1 point, R3=(1+1)÷2=1 point;

[0108] Total risk score: R = 0.25 × 1 + 0.4 × 1 + 0.35 × 1 = 1 point;

[0109] Risk level assessment: R < 3 points, classified as a low-risk product (Category A), with full-area dispatch permissions granted.

[0110] Case 2: Conventional side-by-side refrigerator.

[0111] Indicator data: Unit price: 4,500 yuan (3 points), historical transportation damage rate: 0.3% (3 points), packaging protection level: Level 3 (3 points), IP protection level: IP44 (3 points), maintenance cost as a percentage of total cost: 20% (3 points), long-distance transportation defect rate: 0.5% (3 points).

[0112] Dimensional scoring: R1=3 points, R2=(3+3+3)÷3=3 points, R3=(3+3)÷2=3 points;

[0113] Total risk score: R = 0.25 × 3 + 0.4 × 3 + 0.35 × 3 = 3 points;

[0114] Risk level assessment: 3≤R<5 points, which is a medium-risk product in category B. The amount of goods transferred across regions at one time shall not exceed 30% of the average monthly outbound volume.

[0115] Case 3: Precision Embedded Central Air Conditioner.

[0116] Indicator data: Unit price: 12,000 yuan (7 points), historical transportation damage rate: 1.2% (7 points), packaging protection level: Level 2 (5 points), IP protection level: IP34 (5 points), maintenance cost as a percentage of total cost: 70% (7 points), long-distance transportation defect rate: 1.8% (7 points).

[0117] Dimensional scores: R1=7 points, R2=(7+5+5)÷3≈5.67 points, R3=(7+7)÷2=7 points;

[0118] Total risk score: R = 0.25 × 7 + 0.4 × 5.67 + 0.35 × 7 ≈ 6.52 points

[0119] Risk level assessment: R ≥ 5 points, classified as a Class C high-risk product, with cross-regional transfer permissions locked, and can only be consumed within this region.

[0120] Risk level and scheduling priority are linked: Under the same backlog level, high-risk products have higher scheduling priority than low-risk products, and high-risk backlog inventory is cleared first to minimize product loss risk.

[0121] In this embodiment of the invention, S4 includes performing optimal scheduling execution, which includes a first priority, a second priority, and a third priority.

[0122] In this embodiment of the invention, the first priority is specifically: using the product's region as the scheduling region.

[0123] In this embodiment of the invention, the second optimization stage specifically involves determining edge weights based on an undirected graph model, determining the objective function based on the edge weights, and solving the problem using an improved Dijkstra path optimization method to obtain the scheduling route.

[0124] edge weight The expression is:

[0125] ;

[0126] in, The weighting coefficient for the standardized unit transportation cost. The weighting coefficients for the standardized cost of breakage risk. The weighting coefficient for standardized transportation time costs. The standardized unit transportation cost, For the standardized cost of breakage risk, For standardized transportation time costs, For the undirected graph model, the edges correspond to the transportation routes between warehouses i and j;

[0127] Standardized unit transportation cost, range of values , ;

[0128] Standardized breakage risk cost, range of values , ;

[0129] Standardized transportation time cost, range of values , .

[0130] The expression for the objective function is: ,in, This represents the total weight of the path. For path.

[0131] That is, among all feasible paths from the source warehouse to the target warehouse that satisfy the SKU inventory constraints and transportation risk level restrictions of the transit warehouse, solve for the path. This minimizes the cumulative edge weight of all undirected edges within the path.

[0132] Constraints: Path None of the intermediate warehouses along the route had inventory backlog of this SKU, and the route... Meet the transportation restrictions corresponding to the product's risk level.

[0133] Preliminary constraints: First, supply and demand pairs that meet the conditions for cross-regional scheduling are selected through backlog level matching and risk rule verification;

[0134] Model quantization: Transform areas and transportation routes into nodes and edges of an undirected graph, and transform transportation costs, damage risks, and time costs into computable edge weights;

[0135] Let the undirected graph be As shown in Table 3.

[0136] Table 3

[0137] By improving the shortest path algorithm, the scheduling scheme with the lowest overall cost is found under the premise of satisfying all constraints, so as to achieve the synergistic goal of "backlog reduction" and "cost optimization".

[0138] Taking a typical side-by-side refrigerator (a medium-risk product in Category B) with a backlog of Level 2 inventory as an example, the calculation process for cross-regional dispatch is as follows:

[0139] Step 1: Pre-constraint verification.

[0140] Source area (Sichuan): The refrigerator inventory backlog level is level two, with an average monthly outflow of 100 units;

[0141] Target region (Zhejiang): Refrigerator demand gap rate of 20% (≥15%, meeting the backlog matching conditions);

[0142] Product risk level: Category B, planned inventory transfer quantity is 25 units (≤0.3×100=30 units, which meets the risk verification conditions).

[0143] Step 2: Calculate the weights of the undirected graph model.

[0144] Sichuan ( ) to Zhejiang ( There are two possible routes:

[0145] direct route :

[0146] Transportation cost: 120 yuan / unit, after standardization ;

[0147] Breakage rate: 0.4%, after standardization ;

[0148] Transportation time: 48 hours, after standardization ;

[0149] Overall weighting: .

[0150] transit route (Sichuan → Hubei → Zhejiang) ):

[0151] Weight: ;

[0152] Weight: ;

[0153] Total path weight: .

[0154] Step 3: Solve the algorithm.

[0155] After improving Dijkstra's algorithm, the direct route is calculated. The total weight (0.69) is less than that of the transit route (0.76), so the direct route is determined to be the cost-optimal path, and the dispatch instruction is output: 25 refrigerators are directly transferred from the Sichuan area to the Zhejiang area.

[0156] In this embodiment of the invention, the third priority is specifically: generating a clearing list and sending it to the operation end for processing.

[0157] The following description is based on specific embodiments.

[0158] This embodiment applies to a large-scale electrical appliance distribution company operating across multiple regions in China. Its business covers 6 provincial-level operating areas and 21 regional warehouses, primarily dealing in large appliances such as refrigerators, washing machines, and air conditioners, as well as related appliance accessories. The business exhibits significant seasonal fluctuations and is heavily influenced by major promotional events like 618 and Double 11, resulting in large fluctuations in monthly inventory turnover data. Traditional single-indicator judgment methods are prone to misjudging inventory backlogs. This embodiment employs the technical solution of this invention to achieve accurate identification and intelligent cross-regional scheduling of inventory backlogs. The specific implementation process is as follows:

[0159] Step 1: Collection and preprocessing of inventory data across the entire region.

[0160] We built an enterprise-level central control center and inventory data platform, connecting the WMS warehouse management system, ERP order management system, and TMS transportation management system of all 21 warehouses, and completed the collection and standardized preprocessing of all business data.

[0161] The core data collection dimensions include: beginning / end inventory quantity of a single product, beginning / end inventory balance of a single product, real-time inventory quantity of a single product, inbound time of a single product, inventory age segment data, historical monthly turnover rate data for the past 6 months, monthly sales outbound quantity, monthly sales cost, product unit price, product risk level, warehouse region, inter-warehouse transportation cost and timeliness, and order demand and inventory gap data.

[0162] Preprocessing: Standardize data format and statistical standards, remove abnormal inventory data caused by abnormal orders and inventory discrepancies, fill in missing fields, and classify and store data by region, product category, and risk level to provide standardized data support for subsequent judgment and scheduling.

[0163] Step 2: Standardized assessment and classification of backlogged inventory.

[0164] Based on the business characteristics of the home appliance industry, enterprises can customize the following judgment rules through the system's built-in dynamic configuration module. The system will then automatically complete the identification, verification, and classification of all inventory backlogs:

[0165] For ordinary value home appliances (unit price below 5,000 yuan, including regular home appliance accessories): the initial screening threshold for backlog is set at 0.8 times / month;

[0166] High-value large home appliances (unit price of 5,000 yuan and above, including large home appliances such as refrigerators, washing machines, and air conditioners): The initial screening threshold for backlog is set at 0.6 times / month;

[0167] The system completes the initial screening of all SKUs and identifies a list of inventory items awaiting review whose monthly turnover rate is lower than the corresponding initial screening threshold.

[0168] For the inventory list awaiting review after initial screening, the system performs a two-condition review. If any of the following conditions are met, it is determined to be substantial backlog of inventory, completely avoiding misjudgments caused by monthly data fluctuations due to peak and off-peak seasons:

[0169] Condition 1: The monthly inventory turnover rate of high-value large appliances is lower than the preset backlog threshold of 0.4 times / month for two consecutive calendar months;

[0170] Condition 2: Although the monthly inventory turnover rate is higher than the corresponding initial screening threshold, the inventory age structure has deteriorated, that is, the proportion of inventory aged more than 3 months exceeds 30%, or the proportion of inventory aged more than 6 months exceeds 15%.

[0171] Upon review, the system identified 12 SKUs with substantial overstock inventory during this period.

[0172] Enterprise-defined tiered percentage thresholds: First percentage threshold 5%, second percentage threshold 2%. The system completes backlog tiering and priority sorting according to the following rules:

[0173] Level 1 backlog (extreme backlog): The turnover rate lags behind the judgment threshold by 50% or more, and the inventory capital of a single product accounts for more than 5% of the total inventory capital of the region;

[0174] Level 2 backlog (severe backlog): The turnover rate lags behind the judgment threshold by 30%-50%, and the inventory capital of a single product accounts for more than 2% of the total inventory capital of the region;

[0175] Level 3 backlog (general backlog): Backlogged inventory whose turnover rate lags behind the judgment threshold by less than 30%.

[0176] This period identified 2 Level 1 backlogs, 5 Level 2 backlogs, and 5 Level 3 backlogs. The system automatically tagged all the actual backlogs with a level label, sorted them according to scheduling priority, and pushed them to the central control center's scheduling module.

[0177] Step 3: Configure differentiated control rules based on product risk level.

[0178] Enterprises should configure risk classification and control rules for three categories (A / B / C) based on product transportation loss risk, fragility, and value attributes.

[0179] Category A (Low Risk) Products: Regular household appliance parts, not easily damaged, with low transportation requirements, and are granted normal dispatch permissions across the entire region;

[0180] Category B (Medium Risk) Products: Regular washing machines and refrigerators, which have certain transportation protection requirements. The quantity of goods transferred across regions at one time is limited to no more than 30% of the average monthly outbound volume of the product in that region.

[0181] Category C (High-Risk) Products: Precision central air conditioning units and embedded home appliances are high-value and easily damaged. The system automatically locks cross-regional transfer permissions, allowing them to be consumed only within the current region.

[0182] Configure linkage rules: Under the same backlog level, the scheduling priority of high-risk products is higher than that of low-risk products.

[0183] Step 4: Cost-optimal cross-regional scheduling based on a three-tier priority system. The system strictly adheres to the core principles of "internal first, backlog priority, lowest cost, and controllable risk," executing automated scheduling in a three-tier priority progression:

[0184] First priority: Self-sufficiency within the region. The system prioritizes automatically matching the backlog of inventory within each region with the order demand and inventory gap within that region. In this cycle, a total of 6 SKUs were transferred within the region without the need for cross-regional transfer, minimizing transportation costs. For the 2 high-risk backlog SKUs of Category C that cannot be processed within the region, the system automatically pushes a local rapid clearance reminder to the operations department to execute special promotional clearance.

[0185] Second priority: Cross-regional coordinated scheduling and matching. For the four A / B category backlogged SKUs that cannot be processed within a region, cross-regional coordinated scheduling is triggered. The central control center completes the demand matching across the entire region according to the backlog level from high to low, automatically verifies risk control rules, selects the route with the lowest transportation cost based on the undirected graph scheduling model, generates a transfer order, and issues it for execution after approval by the central control center, thus completing the cross-regional processing of all backlogged SKUs.

[0186] Third priority: Special clearance disposal. For backlogged inventory with no matching demand, the system automatically generates a special clearance list, sorts it by backlog level and risk level, and pushes it to the operations department for channel distribution disposal.

[0187] Step 5: Full-process monitoring, performance evaluation and closed-loop optimization.

[0188] The system provides real-time monitoring of inventory turnover data, backlog clearance progress, transfer cost data, and product loss data across the entire region, automatically generating visual monitoring dashboards and triggering alerts for abnormal data. Inventory turnover rate, backlog clearance rate, transfer cost control, and product loss rate are incorporated into the core performance evaluation indicators for each region. The system reviews the entire business process data monthly, dynamically optimizing backlog judgment thresholds, risk control rules, and scheduling matching logic to continuously adapt to business changes.

[0189] After this implementation, the company's monthly inventory turnover rate for major home appliances increased to 92%, the overall cost of cross-regional inventory transfer decreased by 18%, the transportation loss rate of high-risk products decreased by 32%, and the accuracy rate of identifying backlogged inventory increased to over 95%, effectively avoiding the problem of misjudgment of monthly data fluctuations caused by major promotional activities and peak and off-peak seasons.

[0190] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for resolving inventory backlog and controlling risk of cross-regional scheduling of durable goods, characterized in that, Includes the following steps: S1. Collect the dataset from the entire region and perform preprocessing; S2. Based on the preprocessed dataset, determine the substantial backlog of inventory using the first or second condition, and generate the corresponding scheduling priority. The first condition is set by the monthly inventory turnover rate; The second condition is set by the inventory percentage; S3. Based on the scheduling priority corresponding to the substantial backlog of inventory, determine the product risk level using the total risk score corresponding to the comprehensive score of product value attributes, comprehensive score of transportation vulnerability, and comprehensive score of transportation loss consequences. S4. Based on the product risk level, perform optimal scheduling execution using the first priority, second priority, and third priority; the second priority is obtained by solving the objective function to obtain the scheduling route.

2. The method for resolving inventory backlog and controlling risk of cross-regional scheduling of durable goods according to claim 1, characterized in that, In S1, the dataset includes inventory data, order data, basic product data, and transportation route data.

3. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 1, characterized in that, S2 includes the following sub-steps: S21. Perform preliminary screening of the preprocessed dataset using the basic judgment values; S22. Based on the preliminary screening of the dataset, determine the actual backlog of inventory; S23. Divide the actual backlog of inventory and determine the corresponding scheduling priority.

4. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 3, characterized in that, In step S22, the dataset that meets the first condition or the second condition is treated as substantial backlog inventory. The first condition is specifically: the monthly inventory turnover rate is lower than the preset backlog threshold for the corresponding product category for two consecutive calendar months; The second condition is specifically: the inventory percentage of the preset inventory age segment exceeds the preset percentage threshold.

5. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 3, characterized in that, In S23, the scheduling priority includes first-level backlog, second-level backlog, and third-level backlog; The first-level backlog is specifically defined as: the turnover rate lags behind the judgment threshold by 50% or more, and the inventory capital ratio of a single product exceeds the first ratio threshold of the total inventory capital in the region. The second-level backlog is specifically defined as: the turnover rate lags behind the judgment threshold by 30%-50%, and the proportion of inventory funds for a single product exceeds the second proportion threshold of the total inventory funds in the region. The three-level backlog is specifically defined as a turnover rate that lags behind the judgment threshold by no more than 30%.

6. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 1, characterized in that, In S3, determining the product risk level specifically involves: weighting and fusing the comprehensive score of product value attributes, the comprehensive score of transportation vulnerability, and the comprehensive score of transportation loss consequences to obtain a total risk score; if the total risk score is less than 3, the product risk level is low risk; if the total risk score is greater than or equal to 3 and less than 5, the product risk level is medium risk; if the total risk score is greater than or equal to 5, the product risk level is high risk. The overall risk score The expression is: ； ； ； ； in, The weighting of the overall score for product value attributes. The weighting of the overall score for transport fragility. The weighting of the overall score for the consequences of transportation losses. A comprehensive score for the product's value attributes. For the comprehensive score of transport vulnerability, A comprehensive score is given for the consequences of transportation losses. This refers to the unit purchase price of a single product. Historical transport damage rate For product packaging protection level, For the product's IP protection rating, This represents the percentage of repair costs per unit damaged. For long-distance transportation defect rate.

7. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 1, characterized in that, In S4, optimal scheduling execution includes first priority, second priority, and third priority.

8. The method for resolving inventory backlog and controlling risk of cross-regional dispatching of durable goods according to claim 7, characterized in that, The first priority is specifically: the region to which the product belongs is used as the scheduling region.

9. The method for resolving inventory backlog and controlling risk of cross-regional scheduling of durable goods according to claim 7, characterized in that, The second optimization stage specifically involves: determining edge weights based on the undirected graph model, determining the objective function based on the edge weights, and solving the problem using the improved Dijkstra path optimization method to obtain the scheduling route; The edge weight The expression is: ； in, The weighting coefficient for the standardized unit transportation cost. The weighting coefficients for the standardized cost of breakage risk. The weighting coefficient for standardized transportation time costs. The standardized unit transportation cost, For the standardized cost of breakage risk, For standardized transportation time costs, For the edges of an undirected graph model; The expression for the objective function is: ,in, This represents the total weight of the path. For path.

10. The method for resolving inventory backlog and controlling risk of cross-regional scheduling of durable goods according to claim 7, characterized in that, The third priority is specifically: generating a clearing list and sending it to the operations department for processing.

Images