CN111915254A - An inventory optimization control method and system suitable for auto aftermarket parts - Google Patents

Abstract

The invention provides an inventory optimization control system and method suitable for after-sale auto parts, including: acquiring data on dealers, suppliers and parts, obtaining vehicle ownership data and parts sales summary data by vehicle age; Information; according to the vehicle ownership data of the sub-vehicles, the summary data of parts sales and the engineering information of the parts, establish a forecast management model for the parts sales through the parallel calculation method, and forecast the parts sales; formulate a preset strategy for the designated parts, and manage according to the forecast. The model predicts the sales data of each accessory and calculates the accessory inventory strategy; according to the accessory inventory strategy, the order strategy is calculated for each logistics node by means of timed tasks; the invention realizes the automatic demand prediction of after-sales accessories and the ability to optimize the parameters of the prediction model, which helps after-sales The department saves a lot of time for follow-up processing of invalid information.

Description

An inventory optimization control method and system suitable for auto aftermarket parts

technical field

The invention relates to the field of aftermarket parts supply chain in the automobile industry, in particular to a method and system for inventory optimization control suitable for aftermarket parts of automobiles.

Background technique

As the growth rate of vehicle sales in China slows down, the market has entered the stock market. At the same time, as the profit of vehicle sales decreases year by year, for OEMs and various after-sales service agencies, the proportion of revenue brought by after-sales accessories and related services to their total revenue has increased year by year, and after-sales business has received more and more attention. Compared with vehicle sales, after-sales business management is extensive, especially after-sales parts supply chain management. Inefficient management cannot make the original huge after-sales market reach the expected profit. The problem is mainly manifested in the following aspects:

1) There is a lack of professional after-sales supply chain management tools. Distributors usually use the attached sub-modules of the DMS system. Most of the OEMs rely on some sub-functions of ERP, which cannot fully support the business needs of the after-sales supply chain. After-sales supply chain management mainly relies on senior engineers to obtain data through various data reports, and make decisions accordingly, with relatively low processing efficiency.

2) Due to the huge amount of accessories data, there are few accessories that can be leanly managed. For example, after-sales of a certain brand, its effective accessories catalogue reaches more than 200,000 kinds, and a 4S shop has access to more than 30,000 kinds of accessories every year. The after-sales parts manager handles more than 1,000 kinds of accessories every week. Under normal circumstances, dealers can only manage 10-20% of them effectively, and the scope of lean management of OEMs only accounts for about 40%. Mainly rely on regular audits and report analysis to discover risks and problems, and lack risk control capabilities.

3) Similar management software products, mainly suitable for fast-selling or electronic products, not suitable for the management needs of a single spare part for 10 years and the huge differences in various categories after automobile sales.

Patent document CN111160819A (application number: 201910587197.6) discloses a new vehicle model initial inventory prediction system, the system includes a data acquisition module, a database module and a data analysis and processing module; the data acquisition module is used to input data into the database module; the database module It includes a personal experience database unit, a historical database unit and a dealer database unit. The data analysis and processing module is used to perform big data analysis and data integration on various data in the database module, and predict the initial inventory of parts and components required for new models of cars. The patent document CN111160819A reasonably predicts the initial inventory of various parts and components required for after-sales maintenance of new models of automobiles, and it is not easy to occur that new models of automobiles need maintenance without corresponding parts, so that the inventory of new models of automobile parts and components is not easy to occur. It can meet the needs of after-sales maintenance, reduce the probability of expedited customized production of parts and emergency transportation of parts, reduce emergency transportation costs, and reduce the number of times that affect the normal maintenance of new models of vehicles. This is also a demand forecasting and order optimization method applied to after-sales auto parts, but it is only suitable for parts management at the launch stage of new models of cars, and cannot support parts forecasting and inventory optimization needs in the entire car life cycle.

Patent document CN105389406A (application number: 201410445463.9) discloses a vehicle design reliability evaluation method based on the unit weighted cumulative number of faults, including: acquiring test sample data under at least the first stage in multiple engineering development stages; The sample data respectively calculate the unit weighted cumulative number of faults corresponding to at least the first stage; use the unit weighted cumulative number of faults corresponding to each project development stage to model the gray system to form a grayscale matrix; by solving the grayscale matrix, predict The unit-weighted cumulative number of faults corresponding to the remaining stages in multiple engineering development stages. The patent document CN105389406A predicts the unit weighted cumulative number of faults corresponding to the remaining stages in multiple engineering development stages by solving the grayscale matrix. It can solve the adverse effects of small samples, poor information, complex and unclear system transfer function during the vehicle durability test on the reliability evaluation of the vehicle design, and it is more suitable for quality estimation in the research and development stage, not suitable for automobiles. After-sales field.

SUMMARY OF THE INVENTION

In view of the defects in the prior art, the purpose of the present invention is to provide an inventory optimization control system suitable for after-sales parts of automobiles.

An inventory optimization control system suitable for after-sale auto parts provided according to the present invention includes:

Data management module: obtain dealers, suppliers and accessories data, obtain vehicle ownership data and accessories sales summary data by vehicle age;

Accessory Engineering Module: Get the engineering information of accessories;

Prediction management module: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted;

Strategy management module: formulate a preset strategy for the specified accessories, and calculate the accessories inventory strategy according to the forecast management model to predict the sales data of each accessory;

Supply chain optimization module: According to the spare parts inventory strategy, calculate the order requirements that need to be issued to each supplier every day;

The predictive management model includes building a model for accessory sales to give future sales expectations for accessories.

Preferably, the data management module includes:

The dealer data includes dealer network topology;

the supplier data includes supplier network topology;

The accessory data includes accessory engineering data, accessory change information, accessory sales data and vehicle sales data;

The accessory sales summary data includes feature processing on the accessory data, and the generation includes: historical monthly sales, historical quarterly sales, monthly sales growth rate and quarterly sales growth rate;

The vehicle ownership data by vehicle age includes calculating the distribution number of different vehicles currently owned by the vehicle based on the sales date data of each vehicle, and calculating the monthly total vehicle age by vehicle age at a preset time in the future by using the time series method. The car has data.

Preferably, the spare parts engineering module includes: the engineering information of the spare parts includes: after-sale spare parts holding cost, loss cost and pull strategy; update spare parts service time and scope changes due to supplier change or technology upgrade; maintenance coverage Model information.

Preferably, the forecast management module includes:

Prediction management module M1: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, establish a prediction management model for parts sales through parallel calculation method;

Prediction management module M2: The prediction management model is trained according to the historical sales data of accessories until the prediction error reaches a preset value.

Preferably, the strategy management module includes: outputting the fittings forecasting result from the same regression forecasting management model, using a distribution function to describe the demand for fittings, and calculating the stock level of fittings pipelines according to the business service level, the formula is as follows:

Among them, p is the service level requirement, I is the optimal pipeline inventory to be found, f(x, α, β) is the distribution density function of accessories demand, x represents the demand for accessories, α represents the mathematical expectation of the demand for accessories, and β represents the accessories. The standard deviation of the demand.

Preferably, the supply chain optimization module includes:

Supply chain optimization module M1: Determine the spare parts inventory strategy according to the strategy management module, and calculate the order strategy for each logistics node through timed tasks;

Supply chain optimization module M2: According to the order strategy and inventory strategy, and obtain the inventory quantity and in-transit order data information, calculate the order requirements that need to be issued to each supplier every day.

The order strategy includes: order strategy=inventory strategy value-in-hand inventory value-in-transit order value;

Among them, the on-hand inventory value represents the amount of spare parts available in the warehouse; the order quantity in transit represents the total amount of spare parts for undelivered orders.

Preferably, it also includes a presentation center module and an early warning monitoring module;

The presentation center module includes displaying data results and editing and confirming order results;

The early warning monitoring module includes low inventory warning, delayed arrival warning, seasonal pull warning and abnormal demand warning;

The low inventory warning is triggered when the inventory in hand is lower than the safety inventory;

The arrival delay warning is triggered when the corresponding order has completed the warehousing operation at the expected expiration time;

The seasonal pull warning is triggered when the predicted value of the peak demand for accessories occurs;

The abnormal demand warning is triggered when the dealer's order is greater than the preset value of the historical order.

Preferably, the data information displayed in the presentation center module includes pipeline inventory strategy value, safety stock strategy value, spot orders, node inventory, sales forecast and average sales volume;

Editing and confirming order results in the presentation center module includes: regularly sending order information to the order execution system based on distributed data synchronization technology.

Preferably, the inventory strategy further includes:

According to management requirements, the inventory strategy is limited to a preset range. When the calculated inventory strategy value is within the preset range, the currently calculated inventory strategy value is taken; when the calculated inventory strategy value is not within the preset range, the closest inventory strategy value is taken. Preset range boundary values are output as inventory strategy values.

An inventory optimization control method suitable for after-sale auto parts provided according to the present invention includes:

Data management steps: Obtain dealers, suppliers and accessories data, obtain vehicle ownership data and accessories sales summary data by vehicle age;

Parts engineering steps: get the engineering information of the parts;

Prediction management steps: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted;

Strategy management steps: formulate a preset strategy for the specified accessories, and calculate the accessories inventory strategy according to the forecast management model to predict the sales data of each accessory;

Supply chain optimization steps: According to the spare parts inventory strategy, calculate the order requirements that need to be issued to each supplier every day;

The predictive management model includes building a model for accessory sales to give future sales expectations for accessories.

Compared with the prior art, the present invention has the following beneficial effects:

1. The present invention uses machine learning technology combined with traditional time series statistical principles, makes full use of the sales data accumulated by car companies after sales, and trains the best forecasting model, thereby realizing automatic demand forecasting of after-sales parts and forecasting model parameters optimization capabilities, helping The after-sales department saves a lot of time for follow-up processing of invalid information;

2. The present invention starts from the KPI of the after-sales department and dynamically allocates the capital resources invested by the department in the supply chain by using the operational planning technical data and distributed computing framework, so that the department KPI reaches a local peak within a period of time after each model adjustment. . Combined with the KPI feedback of the actual operation results, the model parameters can be adjusted intelligently, so as to realize the refinement of the supply chain management of online after-sales accessories, help the after-sales department to maximize the use of investment funds, and improve the gross profit rate;

3. The forecast error MAPE of the accessories of the present invention decreases by an average of 9%; the one-time fulfillment rate of the orders from the dealer to the after-sales department reaches 95%; the total inventory amount of the enterprise accessories decreases by 6%, and the peak value decreases by 13%. Provide long-term orders for upstream accessories suppliers It is predicted to improve the on-time supply rate of suppliers and reduce the pressure on their finished product inventory; reduce the data processing work time of the spare parts planning engineer of the after-sales team of the enterprise, by an average of 67%; reduce the sluggish rate and scrap loss of the enterprise's spare parts, and reduce the sluggish rate by 3% .

Description of drawings

Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments with reference to the following drawings:

Fig. 1 is the schematic diagram of this system;

Figure 2 is a flow chart of data processing.

Detailed ways

The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the inventive concept. These all belong to the protection scope of the present invention.

In the after-sales parts supply chain management business scenario of the automotive industry, this system builds a comprehensive machine learning model by analyzing the historical data of after-sales parts sales and the historical data of vehicle sales, calculates the demand distribution estimate for each part, and uses this estimate to combine with After-sales supply chain warehouse management cost, logistics cost, loss cost and other financial indicators make overall planning of spare parts pipeline inventory.

This system can help the after-sales department of automobile enterprises to intelligently manage the demand for all accessories and pipeline inventory, and can optimize the allocation of various after-sales resources under the requirements of the user-specified turnover rate or service satisfaction rate index, so as to improve the gross profit margin and The efficiency of after-sales management personnel.

Example 1

An inventory optimization control system suitable for after-sale auto parts provided according to the present invention includes: as shown in FIG. 1 ,

Data management module: obtain dealers, suppliers and accessories data, obtain vehicle ownership data and accessories sales summary data by vehicle age;

Accessory Engineering Module: Get the engineering information of accessories;

Prediction management module: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted;

Strategy management module: formulate a preset strategy for the specified accessories, and calculate the accessories inventory strategy according to the forecast management model to predict the sales data of each accessory;

Supply chain optimization module: According to the spare parts inventory strategy, calculate the order requirements that need to be issued to each supplier every day;

The predictive management model includes building a model for accessory sales to give future sales expectations for accessories.

Specifically, the data management module includes:

The dealer data includes dealer network topology;

the supplier data includes supplier network topology;

The accessory data includes accessory engineering data, accessory change information, accessory sales data and vehicle sales data;

The accessory sales summary data includes feature processing on the accessory data, and the generation includes: historical monthly sales, historical quarterly sales, monthly sales growth rate and quarterly sales growth rate;

The vehicle ownership data by vehicle age includes calculating the distribution number of different vehicles currently owned by the vehicle based on the sales date data of each vehicle, and calculating the monthly total vehicle age by vehicle age at a preset time in the future by using the time series method. The car has data.

Specifically, the spare parts engineering module includes: the engineering information of the spare parts includes: the holding cost, the loss cost and the pulling strategy of the after-sale spare parts; update the spare parts service time and scope changes caused by the supplier change or technology upgrade; maintenance coverage Model information.

Specifically, the forecast management module includes:

Prediction management module M1: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, establish a prediction management model for parts sales through parallel calculation method;

Prediction management module M2: The prediction management model is trained according to the historical sales data of accessories until the prediction error reaches a preset value.

Specifically, the strategy management module includes: a common regression forecast management model outputs the fitting forecast result, uses a distribution function to describe the fitting demand, and calculates the fitting pipeline inventory level according to the business service level. The formula is as follows:

Among them, p is the service level requirement, I is the optimal pipeline inventory to be found, f(x, α, β) is the distribution density function of the demand for accessories, x is the demand for accessories, α is the mathematical expectation of the demand for accessories, and β is the accessories. The standard deviation of the demand.

Specifically, the supply chain optimization module includes:

Supply chain optimization module M1: Determine the spare parts inventory strategy according to the strategy management module, and calculate the order strategy for each logistics node through timed tasks;

Supply chain optimization module M2: According to the order strategy and inventory strategy, and obtain the inventory quantity and in-transit order data information, calculate the order requirements that need to be issued to each supplier every day.

The order strategy includes: order strategy=inventory strategy value-in-hand inventory value-in-transit order value;

Among them, the on-hand inventory value represents the amount of spare parts available in the warehouse; the order quantity in transit represents the total amount of spare parts for undelivered orders.

Specifically, it also includes a presentation center module and an early warning monitoring module;

The presentation center module includes displaying data results and editing and confirming order results;

The early warning monitoring module includes low inventory warning, delayed arrival warning, seasonal pull warning and abnormal demand warning;

The low inventory warning is triggered when the inventory in hand is lower than the safety inventory;

The arrival delay warning is triggered when the corresponding order has completed the warehousing operation at the expected expiration time;

The seasonal pull warning is triggered when the predicted value of the peak demand for accessories occurs;

The abnormal demand warning is triggered when the dealer's order is greater than the preset value of the historical order.

Specifically, the data information displayed in the presentation center module includes pipeline inventory strategy value, safety stock strategy value, spot orders, node inventory, sales forecast and sales scheduled average;

Editing and confirming order results in the presentation center module includes: regularly sending order information to the order execution system based on distributed data synchronization technology.

Specifically, the inventory strategy further includes:

According to management requirements, the inventory strategy is limited to a preset range. When the calculated inventory strategy value is within the preset range, the currently calculated inventory strategy value is taken; when the calculated inventory strategy value is not within the preset range, the closest inventory strategy value is taken. Preset range boundary values are output as inventory strategy values.

An inventory optimization control method suitable for after-sale auto parts provided according to the present invention includes: as shown in FIG. 1 ,

Data management steps: Obtain dealers, suppliers and accessories data, obtain vehicle ownership data and accessories sales summary data by vehicle age;

Parts engineering steps: get the engineering information of the parts;

Prediction management steps: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted;

Strategy management steps: formulate a preset strategy for the specified accessories, and calculate the accessories inventory strategy according to the forecast management model to predict the sales data of each accessory;

Supply chain optimization steps: According to the spare parts inventory strategy, calculate the order requirements that need to be issued to each supplier every day;

The predictive management model includes building a model for accessory sales to give future sales expectations for accessories.

Specifically, the data management steps include:

The dealer data includes dealer network topology;

the supplier data includes supplier network topology;

The accessory data includes accessory engineering data, accessory change information, accessory sales data and vehicle sales data;

The accessory sales summary data includes feature processing on the accessory data, and the generation includes: historical monthly sales, historical quarterly sales, monthly sales growth rate and quarterly sales growth rate;

The vehicle ownership data by vehicle age includes calculating the distribution number of different vehicles currently owned by the vehicle based on the sales date data of each vehicle, and calculating the monthly total vehicle age by vehicle age at a preset time in the future by using the time series method. The car has data.

Specifically, the parts engineering step includes: the engineering information of the parts includes: after-sale parts holding cost, loss cost and pull strategy; updating parts service time and scope changes due to supplier changes or technology upgrades; maintenance coverage Model information.

Specifically, the forecast management step includes:

Prediction management step M1: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through a parallel computing method;

Prediction management step M2: The prediction management model is trained according to the historical sales data of accessories until the prediction error reaches a preset value.

Specifically, the strategy management step includes: outputting the fittings forecasting result from the same regression forecasting management model, using a distribution function to describe the demand for fittings, and calculating the stock level of fittings and pipes according to the business service level, the formula is as follows:

Among them, p is the service level requirement, I is the optimal pipeline inventory to be found, f(x, α, β) is the distribution density function of accessories demand, x represents the demand for accessories, α represents the mathematical expectation of the demand for accessories, and β represents the accessories. The standard deviation of the demand.

Specifically, the supply chain optimization steps include:

Supply chain optimization step M1: Determine the spare parts inventory strategy according to the strategy management step, and calculate the order strategy for each logistics node through timed tasks;

Supply chain optimization step M2: According to the order strategy and the inventory strategy, and obtain the inventory quantity in hand and the data information of the order in transit, calculate the order requirements that need to be issued to each supplier every day.

The order strategy includes: order strategy=inventory strategy value-in-hand inventory value-in-transit order value;

Among them, the on-hand inventory value represents the amount of spare parts available in the warehouse; the order quantity in transit represents the total amount of spare parts for undelivered orders.

Specifically, it also includes a presentation center step and an early warning monitoring step;

The step of showing the center includes displaying data results and editing and confirming order results;

The early warning monitoring step includes prompting low inventory early warning, delayed arrival early warning, seasonal pull early warning and abnormal demand early warning;

The low inventory warning is triggered when the inventory in hand is lower than the safety inventory;

The arrival delay warning is triggered when the corresponding order has completed the warehousing operation at the expected expiration time;

The seasonal pull warning is triggered when the predicted value of the peak demand for accessories occurs;

The abnormal demand warning is triggered when the dealer's order is greater than the preset value of the historical order.

Specifically, the data information displayed in the step of the presentation center includes pipeline inventory strategy value, safety stock strategy value, spot orders, node inventory, sales forecast and average sales volume;

Editing and confirming the order result in the step of the presentation center includes: regularly sending order information to the order execution system based on the distributed data synchronization technology.

Specifically, the inventory strategy further includes:

According to management requirements, the inventory strategy is limited to a preset range. When the calculated inventory strategy value is within the preset range, the currently calculated inventory strategy value is taken; when the calculated inventory strategy value is not within the preset range, the closest inventory strategy value is taken. Preset range boundary values are output as inventory strategy values.

This system is provided to the after-sales supply chain department of the OEM, and is used to manage the spare parts inventory management of its after-sales spare parts center. The system is responsible for sending orders to its related accessories suppliers according to the calculation strategy every day, and at the same time predicting the after-sales accessories demand of the dealers' stores under its jurisdiction for the center. The system has changed the original state of the enterprise that relies on experience and fixed calculation formulas to predict and optimize inventory, starting from the core indicators: spare parts inventory turnover rate and order fulfillment rate, and reasonably allocate the inventory of each spare part to achieve the best index state.

Through the processing of multiple modules in this system, it can effectively help the after-sales department of the car company to save time and cost, improve the efficiency of personnel in the process of dealing with the parts supply chain, and help the department to formulate the best inventory strategy. Increase after-sale gross profit margin.

Example 2

Example 1 is a variation of Example 2

As shown in Figure 1, the system includes the following seven modules:

The data management module obtains dealer and accessories data from various data interfaces through timed tasks and other methods, and obtains vehicle ownership data and accessories sales summary data by vehicle age, which are used for:

Synchronize parts engineering data, parts change information, dealer network topology, supplier network topology, parts sales data, vehicle sales data and other information;

Perform feature processing on accessory sales data to generate the following indicators: historical monthly sales, historical quarterly sales, monthly sales growth rate, quarterly sales growth and other indicators;

According to the sales date data of each vehicle, the distribution of different vehicle ages of the currently owned vehicles is calculated, and through the time series method, the monthly vehicle ownership by vehicle age in the next 12 months is calculated.

The accessory engineering module, which obtains the engineering information of accessories through file upload and data interface, is used for:

Through file upload, import after-sales proprietary information such as holding cost, loss cost, pull strategy and other after-sales accessories into the system;

By establishing a data interface with the BOM system, update spare parts service time and scope changes due to supplier changes or technology upgrades;

Through the real-time transmission of files, the maintenance strategy of accessories is imported into the system, that is, the information of the model covered by the maintenance.

The forecasting management module, based on the vehicle and accessories sales data obtained and organized by the data management module and the accessories engineering module, models the sales of accessories and gives its future sales expectations, and at the same time automatically adjusts the model parameters according to the forecast error and obtains better results. Predictive effect, which is used to:

Split the historical sales data of accessories into two parts, the training set and the test set. By default, the test set is the sales data of the latest natural month, and the rest is the training set;

According to the characteristics of after-sales data, there are many types of accessories, about 300,000 kinds, and the number of dealers is large, about 1,000. There are huge differences between accessories and accessories and between dealers and dealers, so the model needs to be able to cope with many different characteristics The sales data of , adopt the parallel computing method to design the forecast model, the main steps are as follows:

Each sub-model in the framework outputs the monthly forecast for the next 12 months as required, as well as the monthly forecast and forecast error of the test set. The test set defaults to the latest sales month;

The framework includes time series model and support vector machine model by default, and the framework allows the above;

其中P为测试集于都预测，A为测试集月度实际销量。The framework selects the model with the smallest error as the prediction model according to the monthly relative error of the test set, where the relative error mape formula is,

Among them, P is the prediction of Yudu in the test set, and A is the actual monthly sales of the test set.

The strategy management module provides an entry for system users to input custom strategies for specified accessories, and calculates inventory strategies for each accessory in combination with forecast data. It is used for:

其中p为服务水平要求，I待求最佳管道库存，f(x,α,β)为配件需求的分布密度函数；其中，x表示配件需求量，α表示配件需求量的数学期望，β表示配件需求量的标准差；Output the prediction result of accessories through the prediction module, select a suitable distribution function to describe it, and calculate the inventory level of accessories and pipelines according to the requirements of the business service level.

where p is the service level requirement, I is the optimal pipeline inventory to be found, and f(x,α,β) is the distribution density function of the demand for accessories; where x represents the demand for accessories, α represents the mathematical expectation of the demand for accessories, and β represents the demand for accessories Standard deviation of the demand for accessories;

Maintain the model control limits, limit the calculation results of the inventory strategy model according to management requirements, and give the upper and lower limits of the inventory. Within this range, the system will train and adjust the model parameters according to the data. If the calculated pipeline inventory value exceeds the limit range, take the closest limit boundary value as the pipeline inventory output result;

Customize the accessory strategy formula to set the upper and lower limits of the inventory, and submit the system calculation in the form of Json, where Json refers to the JavaScript object notation, which is a lightweight data exchange format, for example, System:6*D _mean ≤I ≤8*D _mean &Part.Gcc in(BJ K), it means that the spare parts inventory strategy whose Gcc attribute is B, J or K is maintained between 6 and 8 times the weekly average demand, when the system strategy calculation result is between 6 and 8 The calculation result of the system strategy is used between the weekly average demand of the week. If it is less than 6 times the weekly demand, the 6 times the weekly demand is used as the inventory strategy, and if it is greater than 8 times the weekly demand, the 8 times the weekly demand is used as the inventory strategy. Among them, System represents the intersection result between the input strategy and the system calculation, I represents the value of the inventory strategy, D _mean represents the weekly average demand for parts, part.Gcc represents the Gcc attribute of the part, and the Gcc attribute describes the volume and shape of the part, in( BJK) means that the part.gcc attribute is required to be one of B, J, and K, and when the volume and shape parameters of the part are one of B, J, and K, the input strategy takes effect.

Supply Chain Optimization Module, which is used to:

According to the spare parts inventory strategy determined by the strategy management module, the order strategy is calculated for each logistics node by means of timed tasks. Order strategy = inventory strategy value - on-hand inventory value - in-transit order value, where on-hand inventory value refers to the inventory value in the warehouse Amount of spare parts available, in-transit order value refers to the total amount of spare parts for an undelivered order;

According to the inventory strategy and order strategy, and establish a data interface with the logistics warehousing system to obtain information on the number of spare parts in stock and the number of orders in transit, and calculate the order requirements that need to be issued to each supplier every day.

Presentation center module, which is used to:

Display data results, including: pipeline inventory strategy value, safety stock strategy value, spot orders, node inventory, sales forecast, average sales, etc.;

Edit and confirm order results, and regularly send order information to the order execution system based on distributed data synchronization technology.

Early warning monitoring module, which is used for:

Prompt low inventory warning, triggered when the inventory in hand is lower than the safety inventory;

Arrival delay warning, which is triggered when the corresponding order has completed the warehousing operation at the expected expiration time;

Seasonal pull warning, triggered when the peak forecast value of accessories demand occurs;

Abnormal demand warning, triggered when the dealer's order is greater than the thousandth of the historical order.

As shown in Figure 2, it is the data processing flow of this system, including the following steps:

Demand forecast stage:

Training data: The system divides the historical sales data of accessories into a training set and a validation set in chronological order. By default, the validation set is the sales data of the last month, and the rest are training sets;

Model framework: distribute and parallel multiple forecasting models, try multiple modeling methods for each accessory data, the system uses time series model and support vector machine model by default;

Model parameter setting: Use the training set and validation set data to train each available model in the model framework, and calculate the model parameters and error estimates respectively;

Error data: the actual prediction error data of each model accumulated during the system operation;

Model selection: select the model with the smallest average of model error estimates and historical error data as the current prediction model;

Prediction calculation: According to the results selected by the prediction model, the calculation process is performed to obtain the 12-month prediction data of the accessories.

Inventory optimization stage:

Calculation of the upper limit of the distribution model: According to the random inventory model of the operational inventory theory, combined with the forecast calculation results, calculate the upper limit of the inventory of each accessory;

Custom inventory strategy: user-defined inventory requirements input to the system through Json;

Inventory data: Through the data interface with the logistics warehousing system TMS, the spare parts in the warehouse and in-transit data obtained;

Order calculation: Combining II.A, II, B, II.C, and according to the supplier's order requirements uploaded by the user, such as packaging and minimum order quantity, etc., calculate the spare parts order issued to the supplier on the same day;

Order approval: The user confirms the order data calculated by the model in the system, and calls the relevant interface of the order execution system and the enterprise operation management system to send the order.

Those skilled in the art know that, in addition to implementing the system, device and each module provided by the present invention in the form of pure computer readable program code, the system, device and each module provided by the present invention can be completely implemented by logically programming the method steps. The same program is implemented in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, and embedded microcontrollers, among others. Therefore, the system, device and each module provided by the present invention can be regarded as a kind of hardware component, and the modules used for realizing various programs included in it can also be regarded as the structure in the hardware component; A module for realizing various functions can be regarded as either a software program for realizing a method or a structure within a hardware component.

Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above-mentioned specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essential content of the present invention. The embodiments of the present application and features in the embodiments may be combined with each other arbitrarily, provided that there is no conflict.

Claims (10)

1. an inventory optimization control system applicable to auto aftermarket parts, is characterized in that, comprises: Data management module: Obtain dealers, suppliers and accessories data from the network or manual input through the computer, and obtain the vehicle ownership data and accessories sales summary data by vehicle age; Accessory engineering module: Obtain the engineering information of accessories through file upload and data interface; Prediction management module: According to the vehicle ownership data by vehicle age, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted; Strategy management module: formulate a preset strategy for the specified accessories, predict the sales data of each accessory according to the forecast management model, calculate the accessories inventory strategy, and store it in the accessories inventory strategy database; Supply chain optimization module: Obtain the spare parts inventory strategy from the spare parts inventory strategy database, and calculate the order requirements that need to be issued to each supplier every day; The predictive management model includes building a model for accessory sales to give future sales expectations for accessories. 2. The inventory optimization control system suitable for auto aftermarket parts according to claim 1, wherein the data management module comprises: The dealer data includes dealer network topology; the supplier data includes supplier network topology; The accessory data includes accessory engineering data, accessory change information, accessory sales data and vehicle sales data; The accessory sales summary data includes feature processing on the accessory data, and the generation includes: historical monthly sales, historical quarterly sales, monthly sales growth rate and quarterly sales growth rate; The vehicle ownership data by vehicle age includes calculating the distribution number of different vehicles currently owned by the vehicle based on the sales date data of each vehicle, and calculating the monthly total vehicle age by vehicle age at a preset time in the future by using the time series method. The car has data. 3 . The inventory optimization control system suitable for after-sale auto parts according to claim 1 , wherein the parts engineering module comprises: the engineering information of the parts includes: the holding cost, the loss cost and the pull of the after-sale parts. 4 . Policy; update spare parts service hours and scope changes due to supplier changes or technology upgrades; repair coverage model information. 4. The inventory optimization control system suitable for auto aftermarket parts according to claim 1, wherein the forecast management module comprises: Prediction management module M1: According to the vehicle ownership data of sub-vehicles, the summary data of parts sales and the engineering information of parts, establish a prediction management model for parts sales through parallel calculation method; Prediction management module M2: The prediction management model is trained according to the historical sales data of accessories until the prediction error reaches a preset value. 5. The inventory optimization control system suitable for after-sale auto parts according to claim 1, characterized in that, the strategy management module comprises: a synergistic prediction management model outputs parts prediction results, uses a distribution function to describe the demand for parts, and Calculate the inventory level of fittings and pipes according to the business service level, the formula is as follows:

Among them, p is the service level requirement, I is the optimal pipeline inventory to be found, f(x, α, β) is the distribution density function of accessories demand, x represents the demand for accessories, α represents the mathematical expectation of the demand for accessories, and β represents the accessories. The standard deviation of the demand. 6. The inventory optimization control system suitable for auto aftermarket parts according to claim 1, wherein the supply chain optimization module comprises: Supply chain optimization module M1: Determine the spare parts inventory strategy according to the strategy management module, and calculate the order strategy for each logistics node through timed tasks; Supply chain optimization module M2: According to the order strategy and inventory strategy, and obtain the inventory quantity and in-transit order data information, calculate the order requirements that need to be issued to each supplier every day; The order strategy includes: order strategy=inventory strategy value-in-hand inventory value-in-transit order value; Among them, the on-hand inventory value represents the amount of spare parts available in the warehouse; the order quantity in transit represents the total amount of spare parts for undelivered orders. 7. The inventory optimization control system suitable for after-sale auto parts according to claim 1, characterized in that, further comprising a display center module and an early warning monitoring module; The presentation center module includes displaying data results and editing and confirming order results; The early warning monitoring module includes low inventory warning, delayed arrival warning, seasonal pull warning and abnormal demand warning; The low inventory warning is triggered when the inventory in hand is lower than the safety inventory; The arrival delay warning is triggered when the corresponding order has completed the warehousing operation at the expected expiration time; The seasonal pull warning is triggered when the predicted value of the peak demand for accessories occurs; The abnormal demand warning is triggered when the dealer's order is greater than the preset value of the historical order. 8 . The inventory optimization control system suitable for after-sale auto parts according to claim 7 , wherein the data information displayed in the display center module includes pipeline inventory strategy value, safety stock strategy value, spot order, and node inventory. 9 . , sales forecast and average sales volume; Editing and confirming order results in the presentation center module includes: regularly sending order information to the order execution system based on distributed data synchronization technology. 9. The inventory optimization control system suitable for auto aftermarket parts according to claim 5, the inventory strategy further comprises: According to management requirements, the inventory strategy is limited to a preset range. When the calculated inventory strategy value is within the preset range, the currently calculated inventory strategy value is taken; when the calculated inventory strategy value is not within the preset range, the closest inventory strategy value is taken. Preset range boundary values are output as inventory strategy values. 10. An inventory optimization control method suitable for auto aftermarket parts, characterized in that it comprises: Data management steps: Obtain dealers, suppliers and accessories data from the network or manual input through the computer, and obtain the vehicle ownership data and accessories sales summary data by vehicle age; Accessory engineering steps: Obtain the engineering information of accessories through file upload and data interface; Prediction management steps: According to the vehicle ownership data by vehicle age, the summary data of parts sales and the engineering information of parts, a prediction management model is established for parts sales through the parallel calculation method, and the parts sales are predicted; Strategy management steps: formulate a preset strategy for the specified accessories, predict the sales data of each accessory according to the forecast management model, calculate the accessories inventory strategy, and store it in the accessories inventory strategy database; Supply chain optimization steps: Obtain the spare parts inventory strategy from the spare parts inventory strategy database, and calculate the order requirements that need to be issued to each supplier every day; The predictive management model includes building a model for accessory sales to give future sales expectations for accessories.

Images