CN-121998553-A - Warehouse purchase, sale and stock management system and method based on big data analysis

CN121998553ACN 121998553 ACN121998553 ACN 121998553ACN-121998553-A

Abstract

The invention discloses a warehouse purchase, sale and storage management system and method based on big data analysis, which relate to the technical field of library purchase, sale and storage management and comprise the steps of collecting historical borrowing data, reader attribute data and time period data in a library service system, and constructing a book circulation characteristic parameter set based on the historical borrowing data; the method comprises the steps of combining reader attribute data and time period data, carrying out weighted analysis on book circulation characteristic parameters to obtain a comprehensive demand index reflecting book demand intensity, constructing a demand prediction model based on the comprehensive demand index, calculating predicted demand of each book in different time periods, calculating corresponding safety stock quantity according to the predicted demand, generating a replenishment instruction and replenishment quantity when the current stock quantity is lower than the safety stock quantity, and executing the replenishment instruction. The invention effectively reduces the occurrence probability of open books and low-demand book stock backlog.

Inventors

LU RUIJUN
HUANG XUEYING
ZHANG YINYING
LU RUIQIANG

Assignees

广州道宽智能科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. A warehouse purchase, sale and stock management method based on big data analysis is characterized by comprising the following steps: step S1, historical borrowing data, reader attribute data and time period data in a library service system are collected, and data cleaning, duplication removal and standardization processing are carried out; step S2, based on the historical borrowing data, respectively calculating the borrowing frequency, average borrowing duration, reserved number of people and inventory borrowing rate of each book in a preset statistical period, and constructing a book circulation characteristic parameter set; Step S3, combining the reader attribute data with the time period data, and carrying out weighted analysis on the book circulation characteristic parameters to obtain a comprehensive demand index reflecting the book demand intensity; S4, constructing a demand prediction model based on the comprehensive demand index, and calculating the predicted demand of each book in different time periods; s5, calculating the corresponding safety stock quantity according to the predicted demand quantity, and generating a replenishment instruction and replenishment quantity when the current stock quantity is lower than the safety stock quantity; and S6, executing the replenishment instruction.
2. The warehouse entry, sales and storage management method based on big data analysis of claim 1, wherein in step S1, history borrowing data, reader attribute data and time period data in a library service system are collected, specifically: Connecting with the library service system based on a reserved interface; the history borrowing data comprises a book unique identifier, borrowing time, returning time, renewing times, reservation records and inventory change records; The reader attribute data comprise age groups, grades, specialty categories and historical borrowing preference labels of readers; the time period data comprises natural time information, learning period start-stop time, examination period and holiday identification information.
3. The warehouse entry, sales and inventory management method based on big data analysis of claim 2, wherein the data cleaning, deduplication and standardization process comprises: S1-1, eliminating abnormal borrowing records and repeated records in the history borrowing data; S1-2, supplementing the missing fields in the history borrowing data in a history mean value interpolation mode; And step S1-3, carrying out normalization processing on the data with different dimensions for unified modeling.
4. The warehouse entry, sales and storage management method based on big data analysis of claim 3, wherein in step S2, based on the historical borrowing data, respectively calculating borrowing frequency, average borrowing duration, reserved number of people and inventory borrowing rate of each book in a preset statistical period, constructing a book circulation characteristic parameter set, and combining the reader attribute data and time period data, carrying out weighted analysis on the book circulation characteristic parameter to obtain a comprehensive demand index reflecting book demand intensity, wherein the method comprises the following steps: s2-1, calculating circulation characteristic parameters for each book and each book class in a preset statistical period T, wherein the circulation characteristic parameters comprise the total number f of books borrowed in the preset statistical period T, the average value T of the duration of each book borrowing, the number r of readers reserved for the books and the inventory borrowing rate b; Step S2-2, uniformly representing the demand intensity of books by taking the circulation characteristic parameters as a substrate, and constructing a comprehensive demand index D, wherein the calculation formula is as follows, D=alpha, f+beta, r+gamma (1/t) +delta, b; The method comprises the steps of sequentially adding a plurality of weight coefficients, wherein alpha, beta, gamma and delta are weight coefficients of borrowing frequency, average borrowing duration, reserved people and inventory borrowing rate, the weight coefficients are used for adjusting the influence proportion of different flow characteristic parameters in a comprehensive demand index, and the weight coefficients meet the conditions of alpha+beta+gamma+delta=1 through normalization processing.
5. The warehouse entry and sales management method based on big data analysis of claim 4, wherein the method for determining the weight coefficients alpha, beta, gamma and delta is specifically as follows: Step1, acquiring historical borrowing data in N groups of preset statistical periods T as correlation analysis samples, and counting to obtain all books in the preset statistical periods T, wherein for any book a, the total borrowing amount y of the books in the preset statistical periods T, the total borrowing times f a of the books a, the average value T a of each borrowing duration of the books a, the number r a of readers reserving the books a and the inventory borrowing rate b a of the books a are counted; Step2, obtaining the total borrowing amount of books a in N groups of correlation analysis samples, the total borrowing times of the books a, the average value of each borrowing duration of the books a, the number of readers reserved for the books a and the inventory borrowing rate of the books a one by one, wherein the total borrowing amount of the books in the N groups of correlation analysis samples is arranged and recorded as a total borrowing amount sequence Y= (Y 1 ,...,y n ,...,y N ), the borrowing frequency sequence Fa= (f a1 ,...,f an ,...,f aN ) in the N groups of correlation analysis samples is sequentially obtained, the number of readers reserved for the books a sequence Ra= (r a1 ,...,r an ,...,r aN ), the borrowing time reciprocal sequence T 'a= (T' a1 ,...,t' an ,...,t' aN ), the inventory borrowing air rate sequence Ba= (b a1 ,...,b an ,...,b aN ), and Y 1 ,...,y n ,...,y N respectively represents the total borrowing amount of the books reserved for the books a, N represents the total borrowing times of the books a in the N groups of correlation analysis samples, r 3 respectively represents the borrowing frequency sequence Fa= (f a1 ,...,f an ,...,f aN ), N represents the average value of the books a, N represents the reserved for the books a, N represents the average value of the books a, N represents the book in the N numbers of the correlation analysis samples, and N represents the average value of the books in the N groups of the correlation analysis samples, and N represents the average value of the books in the books a 1; Step3, according to the sequences Y, fa, ra, T 'a and Ba, respectively calculating pearson correlation coefficients of the sequences Fa, ra, T' a and Ba and the sequence Y, and respectively marking as rho (Fa, Y), rho (Ra, Y), rho (T 'a, Y) and rho (Ba, Y), wherein rho (Fa, Y) represents pearson correlation coefficients of the borrowing frequency and the total borrowing amount of the book a, rho (Ra, Y) represents pearson correlation coefficients of the number of readers reserving the book a and the total borrowing amount, rho (T' a, Y) represents pearson correlation coefficients of the inverse of the duration of each borrowing of the book a and the total borrowing amount, and rho (Ba, Y) represents pearson correlation coefficients of the inventory borrowing rate and the total borrowing amount of the book a; Step4, based on the obtained pearson correlation coefficient, carrying out standardization, taking the value of rho (Fa, Y) which is completed in standardization as a weight coefficient of the borrowing frequency of the book a, taking the value of rho (Ra, Y) which is completed in standardization as a weight coefficient of the average borrowing duration of the book a, taking the value of rho (T' a, Y) which is completed in standardization as a weight coefficient of the reservation number of the book a, and taking the value of rho (Ba, Y) which is completed in standardization as a weight coefficient of the inventory borrowing rate of the book a.
6. The method for managing inventory sales of a warehouse based on big data analysis of claim 4, wherein the inventory empty rate b is a ratio of an accumulated time length of which the number of books in a preset statistical period T is zero to the preset statistical period T.
7. The warehouse entry, sales and storage management method based on big data analysis of claim 6, wherein in step S4, a demand prediction model is built based on the comprehensive demand index, and the predicted demand of each book in different time periods is calculated, specifically: Step S4-1, taking the D as a unified quantization index for representing the book borrowing demand intensity based on the comprehensive demand index D, and constructing a historical demand index sequence in a time sequence, wherein D is = { D1, the term, dm, the term and DM }, D1, the term, dm, the term and DM respectively represent the comprehensive demand indexes of corresponding books in M preset statistical periods T; Step S4-2, based on a moving average prediction, inputting the historical demand index sequence D = { D1, & gt, dm }, and taking an average value of the integrated demand indexes in the last k statistical periods as a prediction base value of the next period The characteristic is as follows: Wherein k represents the length of the sliding window, and k is a positive integer; Step S4-3, determining a periodic correction factor C according to the average demand index change proportion in the same historical time period, wherein the periodic correction factor C is specifically as follows: wherein, the method comprises the steps of, Representing average integrated demand indexes in the same learning period and holiday type period in the history; representing the average integrated demand index over all statistical periods; step S4-4, introducing a periodic correction factor C, and predicting the comprehensive demand index of the next statistical period Expressed as: 。
8. The warehouse entry and sales management method based on big data analysis of claim 7, wherein in step S5, according to the predicted demand, a corresponding safety stock quantity is calculated, and when the current stock quantity is lower than the safety stock quantity, a replenishment instruction and a replenishment quantity are generated, specifically: S5-1, obtaining a comprehensive demand index predicted value based on prediction Conversion into the predicted borrowing demand μ is characterized by: wherein α and β are the conversion slope and intercept determined by the fit relationship between the historical integrated demand index and the actual borrowing amount; S5-2, calculating the corresponding safety stock quantity S based on statistical distribution, wherein the quantity S is characterized in that S=mu+h is sigma, wherein sigma is the standard deviation of borrowing demand corresponding to the comprehensive demand index of books in a statistical period; And S5-3, if the current stock quantity is lower than the safety stock quantity S, generating a replenishment instruction and replenishment quantity.
9. The method for warehouse entry and sales management based on big data analysis of claim 8, wherein in step S6, executing the replenishment instruction further comprises: And storing and updating the historical borrowing data after one statistical period is finished, re-fitting the fitting relation between the historical comprehensive demand index and the actual borrowing amount, and updating the conversion slope and the intercept.
10. The warehouse entry and sales management system based on big data analysis is characterized by comprising a data acquisition and interface access module, a data preprocessing module, a book circulation characteristic extraction module, a weight self-adaptive calculation module, a comprehensive demand index calculation module, a demand prediction and period correction module and a decision execution and data update module; The data acquisition and interface access module is used for connecting with the library service system based on the reserved interface and acquiring historical borrowing data, reader attribute data and time period data in the library service system; the data preprocessing module is used for performing data cleaning, de-duplication and standardization processing; The book circulation feature extraction module is used for respectively calculating the borrowing frequency, average borrowing duration, reserved number and inventory borrowing rate of each book in a preset statistical period based on historical borrowing data, and constructing a book circulation feature parameter set; The weight self-adaptive calculation module is used for calculating the weight of the circulation characteristic parameters based on a Pearson correlation coefficient method; The comprehensive demand index calculation module is used for carrying out weighted analysis on the book circulation characteristic parameters by combining the reader attribute data and the time period data to obtain a comprehensive demand index reflecting the book demand intensity; The demand prediction and period correction module is used for constructing a demand prediction model based on the comprehensive demand index and calculating the predicted demand of each book in different time periods; The decision execution is used for calculating the corresponding safety stock quantity according to the predicted demand, and generating a replenishment instruction and a replenishment quantity when the current stock quantity is lower than the safety stock quantity The data updating module is used for executing the replenishment instruction and updating the historical borrowing data.

Description

Warehouse purchase, sale and stock management system and method based on big data analysis Technical Field The invention relates to the technical field of library purchase, sale and stock management, in particular to a warehouse purchase, sale and stock management method based on big data analysis. Background With the popularization of library informatization systems, book borrowing, reservation and inventory data can be continuously collected and stored, but the existing book purchase, sale and storage management mode still mainly depends on manual experience or single statistical indexes to make decisions. For example, book purchasing and supplementing are performed only according to the historical borrowing times or the fixed inventory threshold value, so that the real demand condition of books is difficult to comprehensively reflect. In practical application, book borrowing requirements are commonly affected by a plurality of factors, including borrowing frequency, number of reserved persons, borrowing period length, inventory borrowing condition and the like. The prior art generally does not carry out unified quantitative modeling on the multidimensional circulation characteristics, so that the description of the book demand intensity is not accurate enough, and the problems of long-term supply and short-term demand of hot books and low-demand book inventory backlog easily occur. In addition, book borrowing requirements have obvious time periodic characteristics, and the book requirements have obvious fluctuation in different learning period phases, examination periods and holiday periods. The existing management mode mostly adopts static rules or post-hoc analysis, lacks demand prediction capability based on historical data, and is difficult to guide the dynamic adjustment of the inventory structure in time. Therefore, the prior art still has the defects in the aspects of comprehensive evaluation of book demands, demand change trend prediction and dynamic inventory regulation, and a technical scheme for book sales-in management, which can analyze based on multidimensional borrowing data and combine time period factors, is needed. Disclosure of Invention The invention aims to provide a warehouse purchase, sales and inventory management method based on big data analysis, which aims to solve the problems in the prior art. In order to achieve the aim, the invention provides the technical scheme that the library book dynamic purchase, sale and storage management method based on big data analysis specifically comprises the following steps: step S1, historical borrowing data, reader attribute data and time period data in a library service system are collected, and data cleaning, duplication removal and standardization processing are carried out; step S2, based on the historical borrowing data, respectively calculating the borrowing frequency, average borrowing duration, reserved number of people and inventory borrowing rate of each book in a preset statistical period, and constructing a book circulation characteristic parameter set; Step S3, combining the reader attribute data with the time period data, and carrying out weighted analysis on the book circulation characteristic parameters to obtain a comprehensive demand index reflecting the book demand intensity; S4, constructing a demand prediction model based on the comprehensive demand index, and calculating the predicted demand of each book in different time periods; s5, calculating the corresponding safety stock quantity according to the predicted demand quantity, and generating a replenishment instruction and replenishment quantity when the current stock quantity is lower than the safety stock quantity; and S6, executing the replenishment instruction. Further, in step S1, the history borrowing data, the reader attribute data and the time period data in the library service system are collected, specifically: Connecting with the library service system based on a reserved interface; the history borrowing data comprises a book unique identifier, borrowing time, returning time, renewing times, reservation records and inventory change records; The reader attribute data comprise age groups, grades, specialty categories and historical borrowing preference labels of readers; the time period data comprises natural time information, learning period start-stop time, examination period and holiday identification information. Further, the data cleaning, de-duplication and standardization processing includes: S1-1, eliminating abnormal borrowing records and repeated records in the history borrowing data; S1-2, supplementing the missing fields in the history borrowing data in a history mean value interpolation mode; And step S1-3, carrying out normalization processing on the data with different dimensions for unified modeling. Further, in step S2, based on the historical borrowing data, the borrowing frequency, the average borrowing duration, the reserved number of peo