A digital operation method, system and storage medium based on SaaS application

Abstract

The present invention belongs to the field of digital operation technology, and discloses a digital operation method, system and storage medium based on SaaS application: obtaining the first measured value of the sales influencing factor of any retail store at a preset time point, and when the first measured value exceeds the prediction range, calculating the difference between the first measured value and the prediction range, and adjusting the delivery quantity of all first commodities corresponding to the difference of any retail store according to the difference, and obtaining the adjusted delivery quantity of each first commodity respectively; calculating the total adjusted delivery quantity of any first commodity, and if the total adjusted delivery quantity exceeds the inventory quantity, calculating the distribution ratio of any first commodity of any retail store, and distributing commodities to any retail store based on the distribution ratio and the inventory quantity. The technical solution of the present invention generates reliable and highly accurate commodity distribution information by focusing on the measured values ​​of sales influencing factors, and the amount of calculation is small.

Description

Technical Field

[0001] The present invention belongs to the field of digital operation technology, and specifically relates to a digital operation method, system and storage medium based on SaaS application. Background Art

[0002] SaaS is the abbreviation of Software-as-a-service. It is a software application model with software service hosting as the core. It provides software through the Internet. Service providers deploy application software on their own servers. Users can order the required application service software from service providers through the Internet according to their actual needs. The digital operation method based on SaaS applications can bring users more efficient, flexible and convenient business management methods.

[0003] In the prior art, for example, Chinese patent CN116664052B discloses a global digital operation management method and system based on artificial intelligence, which comprehensively analyzes the best distribution data of each retail outlet based on the environmental factors at the location of each retail outlet, establishes the best distribution matrix of the global retail outlet, generates several distribution matrices, determines the distribution priority of each retail outlet based on the sales data of each retail outlet, calculates the difference index matrix between each distribution matrix and the best distribution matrix of the global retail outlet, obtains the global difference index matrix of each distribution matrix, calculates the global fitness of each distribution matrix, selects the distribution matrix with the highest global fitness, and performs the commodity distribution operation management of each retail outlet in the global area according to this distribution matrix. This method analyzes and judges through an algorithm model to achieve real-time analysis and accurate calculation, but the calculation amount is large, and only environmental factors are considered when generating the distribution matrix. Another example is Chinese patent CN117114584B, which discloses a method for intelligent commodity matching and replenishment, which obtains commodity sales data of stores in a region, and analyzes the sales characteristics of different commodities in the stores to form the commodity sales characteristic data of the stores; determines the order quantity of different commodities for each store based on the commodity sales characteristic data to form the store order quantity data; determines the actual order information based on the store order quantity data, and performs first-matching of commodities for different stores based on the actual order information to form the store first-matching goods data, obtains the real-time goods quantity information of the store, and performs replenishment and allocation analysis in combination with the store first-matching goods data to form replenishment and allocation analysis data, and performs replenishment and allocation for different stores based on the replenishment and allocation analysis data. This method does not take into account the total inventory situation when matching and replenishing commodities. When the total inventory is insufficient, it is impossible to replenish according to the replenishment data of each store.

[0004] Therefore, it is an urgent problem to provide a digital operation method, system and storage medium based on SaaS applications to improve the reliability and accuracy of operation decisions while reducing the amount of calculation. Summary of the invention

[0005] In response to the technical problems raised above, the present invention provides a digital operation method, system and storage medium based on SaaS application.

[0006] In a first aspect, the present invention provides a digital operation method based on SaaS application, the method comprising:

[0007] Step 1: A terminal device of any retail store collects a first measured value of a sales influencing factor of any retail store at a preset time point, and sends the first measured value to the SaaS platform;

[0008] Step 2: After the SaaS platform obtains the first measured value, it determines whether the first measured value exceeds the predicted range. If not, the product is delivered according to the predicted delivery quantity. If it exceeds, it proceeds to step 3.

[0009] Step 3: Calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any retail store according to the difference, and obtain the adjusted delivery quantity of each first commodity respectively;

[0010] Step 4: Calculate the total adjusted delivery quantity of any first commodity in all retail stores, and determine whether the total adjusted delivery quantity exceeds the inventory quantity. If not, deliver the commodity to any retail store according to the adjusted delivery quantity of the first commodity in any retail store. If it exceeds, proceed to step 5.

[0011] Step 5: Calculate the allocation ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the allocation ratio and the inventory quantity.

[0012] Specifically, in step 1, sending the first measured value to the SaaS platform includes:

[0013] Step 111: The terminal device divides the first measured value into M1 data blocks according to a preset rule, and assigns a unique data block identifier to each data block, where M1 is a positive integer greater than 1;

[0014] Step 112: search the communication path table based on the destination address of the SaaS platform, randomly select M2 communication paths, and send M1 data blocks to M2 cloud servers on the M2 communication paths respectively, where M2≤M1;

[0015] Step 113: After the sending is completed, the terminal device sends the data sending information to the first server, wherein the data sending information includes the terminal device information, the data identifier of the first measured value, the data block identifier of each data block, the communication path information corresponding to each data block, and the SaaS platform information;

[0016] Step 114: The SaaS platform asks the first server whether there is any information sent to the SaaS platform from the second preset time point, and if so, obtains the terminal device information, the data identifier of the first measured value, the data block identifier of each data block, and the communication path information corresponding to each data block;

[0017] Step 115 , the SaaS platform obtains M1 data blocks from M2 cloud servers based on the data block identifier of each data block and the communication path information corresponding to each data block, and then obtains the first measured value according to the M1 data block using a preset rule.

[0018] Specifically, the process before step 3 includes:

[0019] Step 311: For any retail store, obtain second sales data and second measured values ​​of sales influencing factors of any commodity at each preset time point in a first time period, wherein the first time period is a time period from a first preset time before the collection time of the first measured value;

[0020] Step 312: Calculate the correlation between any commodity in any retail store and the sales influencing factor based on all second sales data and all second measured values ​​in the first time period;

[0021] Step 313: when the absolute value of the correlation is greater than the first preset value and less than or equal to the second preset value, the influence is determined to be level one; when the absolute value of the correlation is greater than the second preset value and less than or equal to the third preset value, the influence is determined to be level two; when the absolute value of the correlation is greater than the third preset value, the influence is determined to be level three;

[0022] Step 314: Store the store identification of any retail store, the product identification of any product, the sales influencing factor, the relevance and the influence in the first data table correspondingly.

[0023] Specifically, step 3 includes:

[0024] Obtaining the absolute value of the difference, and if the absolute value of the difference is less than or equal to the fourth preset value, traversing the first data table, and adjusting the delivery quantity of the product with the third level of influence;

[0025] If the absolute value of the difference is greater than the fourth preset value and less than or equal to the fifth preset value, determine whether the delivery quantity of the commodity with the third degree of influence has been adjusted before the collection time; if not, traverse the first data table to adjust the delivery quantity of the commodities with the third and second degrees of influence; if yes, traverse the first data table to adjust the delivery quantity of the commodity with the second degree of influence;

[0026] If the absolute value of the difference is greater than the fifth preset value, determine whether the delivery quantity of the goods with the third and / or second degree of influence has been adjusted before the collection time. If not, traverse the first data table to adjust the delivery quantity of the goods with the third, second and first degree of influence. If yes, traverse the first data table to adjust the delivery quantity of the goods that have not been adjusted.

[0027] Specifically, step 3 includes:

[0028] Step 321, obtaining a sequence of measured values ​​of sales influencing factors at all preset time points on the current date of any retail store;

[0029] Step 322: traverse the second data table based on the measured value sequence, and obtain the date of the first historical measured value sequence with the greatest similarity to the measured value sequence as a reference date, wherein the second data table is a table storing the historical measured values ​​of the sales influencing factors of any retail store, and the collection time of the first historical measured value sequence is the same as the collection time of the measured value sequence;

[0030] Step 323: Use the historical delivery quantity of any retail store at the next delivery time point corresponding to any first commodity on the reference date as the adjusted delivery quantity of any first commodity in any retail store.

[0031] Specifically, in step 5, the calculation method of the distribution ratio is:

[0032] Get the inventory of any first commodity in any retail store before it opens for business on the day, and calculate the inventory ratio of any first commodity in any retail store. The calculation formula is:

[0033]

[0034] Among them, PI ij is the inventory ratio of the jth first product in the ith retail store, NI ij is the inventory of the jth first product in the ith retail store, i is a positive integer from 1 to I, and I is the total number of retail stores;

[0035] Obtain the total sales volume of any first commodity in any retail store within the second preset time from the material preparation time, and calculate the sales ratio of any first commodity in any retail store. The calculation formula is:

[0036]

[0037] Among them, PS ij is the sales ratio of the jth first product in the i-th retail store, TS ij is the total sales volume of the jth first product in the i-th retail store, i is a positive integer from 1 to I, and I is the total number of retail stores;

[0038] The allocation ratio of any first commodity in any retail store is calculated based on the inventory ratio and sales ratio of any first commodity in any retail store. The calculation formula is:

[0039]

[0040] Among them, D ij is the allocation ratio of the jth first product in the i-th retail store, α i and β i is the parameter weight coefficient.

[0041] Specifically, before step 1, it includes:

[0042] At a predetermined time on the previous day, obtain the predicted values ​​of all supply influencing factors for any commodity in any retail store at each preset time point on the next day, wherein the supply influencing factors at least include sales influencing factors, the commodity number of any commodity, the actual sales volume of any commodity, and the store logo of any retail store;

[0043] Generate a predicted value time series for each supply influencing factor, and then based on the identifier of any supply influencing factor, input the predicted value time series of any supply influencing factor into a feature extraction module corresponding to any supply influencing factor to obtain a first feature vector of the predicted value time series of any supply influencing factor;

[0044] All first feature vectors are input into the first prediction model, and the predicted delivery quantity of any commodity in any retail store at each supply time point is output.

[0045] Specifically, the generation method of the first prediction model is:

[0046] Step 121: randomly extract a second historical measured value sequence of any supply influencing factor of any commodity in one day from the historical distribution volume data table, wherein the historical distribution volume data table at least stores the actual supply volume of any commodity at each supply time point and the measured values ​​of all supply influencing factors at each preset time point;

[0047] Step 122: Based on the identifier of any supply influencing factor, input the second historical measured value sequence of any supply influencing factor into the feature extraction module corresponding to any supply influencing factor to obtain a second feature vector of the second historical measured value sequence of any supply influencing factor;

[0048] Step 123: input all second eigenvectors into a neural network model for training to generate an initial prediction model;

[0049] Step 124: Obtain a third historical measured value sequence different from the second historical measured value sequence from the historical delivery volume data table, generate a third feature vector according to the method of step 122, and then input all third feature vectors into the initial prediction model to obtain the first predicted delivery volume of any commodity at each supply time point;

[0050] Step 125: calculate the deviation between the first predicted delivery quantity and the actual supply quantity at each delivery time point, adjust the parameters of the initial prediction model based on the deviation to generate a new initial prediction model, define the third historical measured value sequence as the second historical measured value sequence, and then repeat steps 124 and 125 until the deviation converges or reaches the preset number of training times, and use the new initial prediction model as the first prediction model.

[0051] In a second aspect, the present invention further provides a digital operation system based on SaaS application, the system comprising: a terminal device and a SaaS platform, the SaaS platform comprising a first judgment module, a first adjustment module, a second judgment module and a second adjustment module;

[0052] The terminal device is used to collect a first measured value of a sales influencing factor of any retail store at a preset time point, and send the first measured value to the SaaS platform;

[0053] A first judgment module is used to judge whether the first measured value exceeds the predicted range. If not, the goods are delivered according to the predicted delivery quantity. If it exceeds, the first adjustment module is entered;

[0054] A first adjustment module is used to calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any retail store according to the difference, and obtain the adjusted delivery quantity of each first commodity respectively;

[0055] The second judgment module is used to calculate the total adjusted delivery quantity of all retail stores for any first commodity, and judge whether the total adjusted delivery quantity exceeds the inventory quantity. If not, the commodity is delivered to any retail store according to the adjusted delivery quantity of the first commodity of any retail store. If it exceeds, the second adjustment module is entered;

[0056] The second adjustment module is used to calculate the allocation ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the allocation ratio and the inventory quantity.

[0057] In a third aspect, the present invention provides a computer storage medium storing program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute any one of the above-mentioned digital operation methods based on SaaS applications.

[0058] The present invention discloses a digital operation method, system and storage medium based on SaaS application, which collects the first measured value of the sales influencing factor of any retail store at a preset time point, determines whether the first measured value exceeds the predicted range, and if not, distributes the goods according to the predicted distribution volume; if it exceeds, calculates the difference between the first measured value and the predicted range, and adjusts the distribution volume of all first goods corresponding to the difference of any retail store according to the difference, obtains the adjusted distribution volume of each first product respectively, and performs regular analysis on the data affecting the sales volume; if the measured value of the sales influencing factor fluctuates greatly compared with the predicted value, adjusts the distribution volume of the goods corresponding to the difference according to the difference, and reduces the number of distribution volume adjustments to be performed simultaneously. The number of commodities is adjusted in stages to reduce the calculation pressure, the total adjusted delivery quantity of all retail stores for any first commodity is calculated, and it is determined whether the total adjusted delivery quantity exceeds the inventory quantity. If not, the commodity is delivered to any retail store according to the adjusted delivery quantity of the first commodity of any retail store. If it exceeds, the allocation ratio of any retail store to any first commodity is calculated, and the commodity is delivered to any retail store based on the allocation ratio and the inventory quantity. When the sales volume of the commodity is large and the inventory is insufficient for allocation, the commodity is allocated according to the real-time sales situation of each retail store to balance the inventory and profit and increase the revenue. According to the technical solution of the present invention, the reliability and accuracy of the operation decision are improved, and the method is simple and the calculation amount is small. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0060] Figure 1 This is a flow chart of a digital operation method based on SaaS application of the present invention;

[0061] Figure 2 It is a structural diagram of a digital operation system based on SaaS application of the present invention. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. Obviously, the specific embodiments described herein are only used to explain the present invention and are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0063] It should be noted that if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0064] Figure 1 FIG. 1 is a flow chart of an embodiment of a digital operation method based on SaaS application provided by the present invention, and the flow chart specifically includes:

[0065] Step 1: A terminal device of any retail store collects a first measured value of a sales influencing factor of any retail store at a preset time point, and sends the first measured value to the SaaS platform.

[0066] The preset time point is set according to the experience of those skilled in the art or according to the actual application scenario, and the embodiment of the present application is not limited to this. For example, the preset time points in a day are 10:00, 12:00, 14:00, and 16:00.

[0067] The sales influencing factors at least include human traffic, air pressure, temperature, rainfall, air humidity, etc. A sales influencing factor data table is established for each retail store at the operation terminal, and the preset collection time points and the measured values ​​of each sales influencing factor at each preset collection time point are correspondingly stored in the sales influencing factor data table of each retail store.

[0068] Since the warehouse capacity of retail stores is limited, in order to ensure that the number of goods in retail stores meets sales demand, multiple goods delivery to retail stores is required during business hours of the day. Regular analysis of the measured values ​​of sales influencing factors and accurate adjustment of the supply volume of retail stores at the next supply time point can reduce inventory pressure and increase profits.

[0069] Specifically, in step 1, sending the first measured value to the SaaS platform includes:

[0070] Step 111: The terminal device divides the first measured value into M1 data blocks according to a preset rule, and assigns a unique data block identifier to each data block, where M1 is a positive integer greater than 1.

[0071] Step 112: search the communication path table based on the destination address of the SaaS platform, randomly select M2 communication paths, and send the M1 data blocks to the M2 cloud servers on the M2 communication paths respectively, where M2≤M1.

[0072] Step 113: After the sending is completed, the terminal device sends the data sending information to the first server, wherein the data sending information includes the terminal device information, the data identifier of the first measured value, the data block identifier of each data block, the communication path information corresponding to each data block, and the SaaS platform information.

[0073] Step 114, the SaaS platform starts to inquire from the first server whether there is any information sent to the SaaS platform from the second preset time point, and if so, obtains the terminal device information, the data identifier of the first measured value, the data block identifier of each data block, and the communication path information corresponding to each data block.

[0074] Step 115 , the SaaS platform obtains M1 data blocks from M2 cloud servers based on the data block identifier of each data block and the communication path information corresponding to each data block, and then obtains the first measured value according to the M1 data block using a preset rule.

[0075] The second preset time point is set according to the experience of those skilled in the art or according to the actual application scenario, and the embodiment of the present application does not limit this. For example, when the preset time point is 10:00, the second preset time point is 10:05.

[0076] Preferably, when dividing the data blocks, the number of divisions each time may be different, that is, the value of M1 is different each time the division is performed.

[0077] The first measured value is divided using the above preset rule, and the first measured value can only be generated when the receiving end receives all data blocks. Preferably, the SaaS platform (i.e., the receiving end) stores the preset rule, and the data sending information includes the preset rule identification information, and the SaaS platform can know the preset rule used by the sending end device based on the identification information.

[0078] When the terminal device sends data, it first divides the data into multiple data blocks, and then sends the above data blocks through different communication paths. When M2 is less than M1, multiple data blocks can be sent on one communication path. Preferably, the cloud server first authenticates the terminal device sending the data, and after the authentication is passed, receives and stores the data blocks sent by the terminal device. After the SaaS platform receives the relevant information of the sent data from the first server, based on the communication path information corresponding to each data block, it searches for the cloud server information (address, etc.) on each communication path, and then obtains the corresponding data block from the cloud server. Preferably, the cloud server first authenticates the terminal device sending the data, and after the authentication is passed, it sends the data block to the SaaS platform.

[0079] Preferably, the data block is deleted after being stored in the cloud server for a third preset time.

[0080] According to the technical solution of the present invention, when sending data, the data is divided into multiple blocks and sent through different communication paths, and the data is cached on different servers, which improves the security of data transmission and avoids the leakage of commercial information.

[0081] Step 2: Determine whether the first measured value exceeds the predicted range. If not, deliver the goods according to the predicted delivery quantity. If it exceeds, proceed to step 3.

[0082] The above prediction range is set according to the predicted value of the sales influencing factors of the previous day. For example, the predicted value of temperature is determined according to the prediction of the weather forecast, and the prediction range is [T-Δ, T+Δ], where Δ is the temperature fluctuation value.

[0083] The first measured value does not exceed the forecast range, indicating that the fluctuation of the measured value of the sales influencing factor is within the allowable range and there is no need to adjust the forecasted delivery volume.

[0084] Preferably, based on the predicted value sequence of the sales influencing factors of any retail store at each preset time point, the second data table is traversed to obtain the date of the historical measured value sequence with the greatest similarity to the above-mentioned predicted value sequence as the predicted reference date, and the historical delivery volume of any retail store at each supply time point corresponding to any commodity on the above-mentioned predicted reference date is used as the predicted delivery volume of any commodity at each supply time point in any retail store. The above-mentioned historical measured value sequence can be the measured value sequence of any of the above-mentioned retail stores, or it can be the measured value sequence of other retail stores. The retail store corresponding to the historical measured value sequence is defined as the selected retail store, and the historical delivery volume of the selected retail store at each supply time point corresponding to any commodity on the above-mentioned predicted reference date is used as the predicted delivery volume of any commodity at any retail store at each supply time point.

[0085] Step 3: Calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any retail store according to the difference, and obtain the adjusted delivery quantity of each first commodity respectively.

[0086] Specifically, the process before step 3 includes:

[0087] Step 311: For any retail store, obtain the second sales data and the second measured value of the sales influencing factor of any commodity at each preset time point in a first time period, wherein the first time period is a time period from a first preset time before the collection time of the first measured value.

[0088] Step 312: Calculate the correlation between any commodity in any retail store and the sales influencing factor based on all the second sales data and all the second measured values ​​in the first time period.

[0089] Step 313: When the absolute value of the correlation is greater than the first preset value and less than or equal to the second preset value, the influence is determined to be level one; when the absolute value of the correlation is greater than the second preset value and less than or equal to the third preset value, the influence is determined to be level two; when the absolute value of the correlation is greater than the third preset value, the influence is determined to be level three.

[0090] Step 314: Store the store identification of any retail store, the product identification of any product, the sales influencing factor, the relevance and the influence in the first data table correspondingly.

[0091] The first preset time is set according to the experience of those skilled in the art or according to the actual application scenario, and the embodiment of the present application does not limit this. Exemplarily, the first preset time is 7 days or 15 days.

[0092] The first preset value, the second preset value and the third preset value are set according to the experience of those skilled in the art or according to actual application scenarios, and the embodiments of the present application are not limited to this.

[0093] The correlation between product sales and sales influencing factors will continue to fluctuate. Based on the measured values ​​of product sales and sales influencing factors in the first time period, the correlation between the two is calculated, and the correlation accuracy is high.

[0094] Products with a third degree of influence are the products most affected by sales influencing factors, products with a second degree of influence are the products second most affected by sales influencing factors, and products with a first degree of influence are the products least affected by sales influencing factors.

[0095] As a preferred technical solution of the present invention, the correlation and influence are updated and stored periodically. When the first measured value exceeds the predicted range, it is determined whether the stored correlation and influence between the sales volume of the product and the measured value of the sales influencing factor are within the validity period. If so, the product delivery quantity is adjusted based on the stored influence. Based on this, the amount of calculation can be reduced and the calculation efficiency can be improved.

[0096] Specifically, step 3 includes:

[0097] The absolute value of the difference is obtained. If the absolute value of the difference is less than or equal to a fourth preset value, the first data table is traversed to adjust the delivery quantity of the products with a third level of influence.

[0098] If the absolute value of the difference is greater than the fourth preset value and less than or equal to the fifth preset value, determine whether the delivery quantity of the goods with a degree of influence of level three has been adjusted before the collection time. If not, traverse the first data table to adjust the delivery quantity of the goods with a degree of influence of level three and level two. If yes, traverse the first data table to adjust the delivery quantity of the goods with a degree of influence of level two.

[0099] If the absolute value of the difference is greater than the fifth preset value, determine whether the delivery quantity of the goods with the third and / or second degree of influence has been adjusted before the collection time. If not, traverse the first data table to adjust the delivery quantity of the goods with the third, second and first degree of influence. If yes, traverse the first data table to adjust the delivery quantity of the goods that have not been adjusted.

[0100] The fourth preset value and the fifth preset value are set according to the experience of those skilled in the art or according to actual application scenarios, and the embodiments of the present application are not limited to this.

[0101] When adjusting the delivery volume of goods, give priority to goods that are most affected by sales-influencing factors. Adjust the sales volume of goods according to the degree of influence of sales-influencing factors. This can reduce the number of goods that need to have their delivery volumes adjusted at the same time, make adjustments in stages, and reduce calculation pressure.

[0102] Specifically, step 3 includes:

[0103] Step 321: Obtain a sequence of measured values ​​of sales influencing factors at all preset time points on the current date of any retail store.

[0104] Step 322: traverse the second data table based on the measured value sequence, and obtain the date of the first historical measured value sequence with the greatest similarity to the measured value sequence as a reference date, wherein the second data table is a table storing the historical measured values ​​of the sales influencing factors of any retail store, and the collection time of the first historical measured value sequence is the same as the collection time of the measured value sequence.

[0105] Step 323: Use the historical delivery quantity of any retail store at the next delivery time point corresponding to any first commodity on the reference date as the adjusted delivery quantity of any first commodity in any retail store.

[0106] Preferably, the above-mentioned first historical measured value sequence can be a measured value sequence of any of the above-mentioned retail stores, or a measured value sequence of other retail stores. The retail store corresponding to the first historical measured value sequence is defined as a selected retail store, and the historical delivery quantity of the selected retail store at each supply time point corresponding to any commodity on the above-mentioned prediction reference date is used as the predicted delivery quantity of any commodity in any retail store at each supply time point.

[0107] Step 4: Calculate the total adjusted delivery quantity of any first commodity in all retail stores, and determine whether the total adjusted delivery quantity exceeds the inventory quantity. If not, deliver the commodity to any retail store according to the adjusted delivery quantity of the first commodity in any retail store. If it exceeds, proceed to step 5.

[0108] For any first commodity, if only some of the retail stores need to adjust the delivery quantity, the total adjusted delivery quantity of any first commodity is the sum of the predicted delivery quantity of the retail stores that do not need to be adjusted and the adjusted delivery quantity of the retail stores that need to be adjusted. If the total adjusted delivery quantity of any first commodity does not exceed the inventory quantity, the retail stores that do not need to be adjusted will be allocated according to the predicted delivery quantity, and the retail stores that need to be adjusted will be allocated according to the adjusted delivery quantity.

[0109] Step 5: Calculate the allocation ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the allocation ratio and the inventory quantity.

[0110] Specifically, in step 5, the calculation method of the distribution ratio is:

[0111] Get the inventory of any first commodity in any retail store before it opens for business on the day, and calculate the inventory ratio of any first commodity in any retail store. The calculation formula is:

[0112]

[0113] Among them, PI ij is the inventory ratio of the jth first product in the ith retail store, NI ij is the inventory of the jth first product in the ith retail store, i is a positive integer from 1 to I, and I is the total number of retail stores;

[0114] Obtain the total sales volume of any first commodity in any retail store within the second preset time from the material preparation time, and calculate the sales ratio of any first commodity in any retail store. The calculation formula is:

[0115]

[0116] Among them, PS ij is the sales ratio of the jth first product in the i-th retail store, TS ij is the total sales volume of the jth first product in the i-th retail store, i is a positive integer from 1 to I, and I is the total number of retail stores;

[0117] The allocation ratio of any first commodity in any retail store is calculated based on the inventory ratio and sales ratio of any first commodity in any retail store. The calculation formula is:

[0118]

[0119] Among them, D ij is the allocation ratio of the jth first product in the i-th retail store, α i and β i is the parameter weight coefficient.

[0120] The second preset time is set according to the experience of those skilled in the art or according to actual application scenarios, and the embodiments of the present application do not limit this.

[0121] The material preparation time is the time to start the distribution of goods to each retail store. The material preparation time and the second preset time are set according to the experience of those skilled in the art or according to the actual application scenario, and the embodiment of the present application does not limit this. For example, the material preparation time is 14:00 and the second preset time is 15 minutes.

[0122] The inventory quantity of any first commodity of any retail store before it starts business on the day is the sum of the inventory at the end of business on the previous day and the quantity of goods delivered after the first delivery on the day.

[0123] Allocating a larger inventory to a retail store indicates that the retail store has a greater sales opportunity according to historical sales records. However, due to uncertain factors, the sales volume of each sales store will fluctuate. Retail stores with large sales in the past will have smaller sales today, while retail stores with small sales in the past will have larger sales today. Therefore, when calculating the allocation ratio of each retail store, the inventory ratio and sales ratio of the product are considered. Preferably, when calculating the inventory ratio, α i >β i .

[0124] The distribution ratio of any first commodity in any retail store is multiplied by the inventory quantity of any first commodity to obtain the commodity distribution quantity of any first commodity in any retail store.

[0125] Specifically, before step 1, it includes:

[0126] At a predetermined time on the previous day, obtain the predicted values ​​of all supply influencing factors for any commodity in any retail store at each preset time point on the next day, wherein the supply influencing factors at least include sales influencing factors, the commodity number of any commodity, the actual sales volume of any commodity, and the store logo of any retail store;

[0127] Generate a predicted value time series for each supply influencing factor, and then based on the identifier of any supply influencing factor, input the predicted value time series of any supply influencing factor into a feature extraction module corresponding to any supply influencing factor to obtain a first feature vector of the predicted value time series of any supply influencing factor;

[0128] All first feature vectors are input into the first prediction model, and the predicted delivery quantity of any commodity in any retail store at each supply time point is output.

[0129] The scheduled time is set according to the experience of those skilled in the art or according to the actual application scenario, and the embodiments of the present application do not limit this. For example, the scheduled time is 23:00.

[0130] Setting a corresponding feature extraction module for each supply influencing factor can improve the accuracy of feature vector extraction.

[0131] Preferably, before inputting the first feature vector into the above-mentioned first prediction model, each first feature vector is weighted, and the weighted first feature vector is input into the above-mentioned first prediction model to predict the delivery quantity of any commodity in any retail store at each delivery time point on the second day to obtain the predicted delivery quantity.

[0132] Specifically, the generation method of the first prediction model is:

[0133] Step 121: randomly extract a second historical measured value sequence of any supply influencing factor of any commodity within one day from the historical distribution volume data table, wherein the historical distribution volume data table at least stores the actual supply volume of any commodity at each supply time point and the measured values ​​of all supply influencing factors at each preset time point.

[0134] Step 122: Based on the identification of any supply influencing factor, input the second historical measured value sequence of any supply influencing factor into the feature extraction module corresponding to any supply influencing factor to obtain a second feature vector of the second historical measured value sequence of any supply influencing factor.

[0135] Step 123: input all second eigenvectors into the neural network model for training to generate an initial prediction model.

[0136] Step 124: Obtain a third historical measured value sequence different from the second historical measured value sequence from the historical distribution volume data table, generate a third eigenvector according to the method of step 122, and then input all third eigenvectors into the initial prediction model to obtain the first predicted distribution volume of any commodity at each supply time point.

[0137] Step 125: calculate the deviation between the first predicted delivery quantity and the actual supply quantity at each delivery time point, adjust the parameters of the initial prediction model based on the deviation to generate a new initial prediction model, define the third historical measured value sequence as the second historical measured value sequence, and then repeat steps 124 and 125 until the deviation converges or reaches the preset number of training times, and use the new initial prediction model as the first prediction model.

[0138] The historical delivery volume data table of the embodiment of the present invention is described by taking Table 1 as an example.

[0139] Table 1

[0140]

[0141] The second historical measured value sequence of the store identification corresponding to the product with product identification 002 is 1, the second historical measured value sequence of the product number is 002, the second historical measured value sequence of the traffic volume is (300, 500, 200, 210, 420, 350), and the second historical measured value sequence of the temperature is (17, 22, 24, 23, 18, 16).

[0142] Preferably, since each retail store is located in a different area and has different types of residents, the first prediction model can be trained for each retail store separately to further improve the accuracy of the model.

[0143] According to the technical solution of the present invention, the accuracy of commodity delivery quantity prediction can be improved and the error can be reduced.

[0144] Figure 2 FIG. 1 is a schematic diagram of a structure of an embodiment of a digital operation system based on SaaS application provided by the present invention. Figure 2 As shown, the system includes: a terminal device 100 and a SaaS platform 200 , and the SaaS platform 200 includes a first judgment module 201 , a first adjustment module 202 , a second judgment module 203 and a second adjustment module 204 .

[0145] The terminal device 100 is used to collect the first measured value of the sales influencing factor of any retail store at a preset time point, and send the first measured value to the SaaS platform.

[0146] The first judgment module 201 is used to judge whether the first measured value exceeds the prediction range. If not, the goods are distributed according to the predicted distribution volume. If it exceeds, the first adjustment module is entered.

[0147] The first adjustment module 202 is used to calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any retail store according to the difference, and obtain the adjusted delivery quantity of each first commodity.

[0148] The second judgment module 203 is used to calculate the total adjusted delivery quantity of all retail stores for any first commodity, and determine whether the total adjusted delivery quantity exceeds the inventory quantity. If not, the commodity is delivered to any retail store according to the adjusted delivery quantity of the first commodity of any retail store. If it exceeds, enter the second adjustment module.

[0149] The second adjustment module 204 is used to calculate the allocation ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the allocation ratio and the inventory quantity.

[0150] According to another aspect of an embodiment of the present invention, a computer storage medium is provided, wherein the computer storage medium stores program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute any one of the above-mentioned digital operation methods based on SaaS applications.

[0151] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0152] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The above-mentioned program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0153] The above-mentioned embodiments only express the preferred implementation mode of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A digital operation method based on SaaS application, characterized in that: The steps include: Step 1: A terminal device of any retail store collects a first measured value of a sales influencing factor of any retail store at a preset time point, and sends the first measured value to the SaaS platform; Step 2: After obtaining the first measured value, the SaaS platform determines whether the first measured value exceeds the predicted range. If not, the product is delivered according to the predicted delivery quantity. If it exceeds, the process proceeds to step 3. Step 3: Calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any of the retail stores according to the difference, and obtain the adjusted delivery quantity of each first commodity respectively; Step 4: Calculate the total adjusted delivery quantity of any first commodity in all retail stores, and determine whether the total adjusted delivery quantity exceeds the inventory quantity. If not, deliver the commodity to any retail store according to the adjusted delivery quantity of any first commodity in any retail store. If it exceeds, proceed to step 5. Step 5: Calculate the distribution ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the distribution ratio and the inventory quantity; The step 3 includes: Step 311: For any retail store, obtain second sales data of any commodity and a second measured value of the sales influencing factor at each preset time point in a first time period, wherein the first time period is a time period of a first preset time before the collection time of the first measured value; Step 312: Calculate the correlation between any commodity in any retail store and the sales influencing factor based on all second sales data and all second measured values ​​in the first time period; Step 313: when the absolute value of the correlation is greater than the first preset value and less than or equal to the second preset value, the influence is determined to be level one; when the absolute value of the correlation is greater than the second preset value and less than or equal to the third preset value, the influence is determined to be level two; when the absolute value of the correlation is greater than the third preset value, the influence is determined to be level three; Step 314: store the store ID of any retail store, the product ID of any product, the sales influencing factor, the relevance, and the influence in a first data table in a corresponding manner; The step 3 comprises: Obtaining the absolute value of the difference, and if the absolute value of the difference is less than or equal to a fourth preset value, traversing the first data table, and adjusting the delivery quantity of the commodity with the third level of influence; If the absolute value of the difference is greater than the fourth preset value and less than or equal to the fifth preset value, determine whether the delivery quantity of the commodity with the third degree of influence has been adjusted before the collection time; if not, traverse the first data table to adjust the delivery quantity of the commodities with the third and second degrees of influence; if yes, traverse the first data table to adjust the delivery quantity of the commodity with the second degree of influence; If the absolute value of the difference is greater than the fifth preset value, determining whether the delivery quantity of the commodities with the third and / or second degree of influence has been adjusted before the collection time, if not, traversing the first data table to adjust the delivery quantity of the commodities with the third, second and first degree of influence, if yes, traversing the first data table to adjust the delivery quantity of the commodities that have not been adjusted; The step 3 comprises: Step 321, obtaining a sequence of measured values ​​of the sales influencing factors at all preset time points on the current date of any retail store; Step 322: traverse the second data table based on the measured value sequence, and obtain the date of the first historical measured value sequence with the greatest similarity to the measured value sequence as a reference date, wherein the second data table is a table storing the historical measured values ​​of the sales influencing factors of any retail store, and the collection time of the first historical measured value sequence is the same as the collection time of the measured value sequence; Step 323: taking the historical delivery quantity of any retail store at the next delivery time point corresponding to any first commodity on the reference date as the adjusted delivery quantity of any first commodity of any retail store; In step 5, the calculation method of the distribution ratio is: Get the inventory of any first commodity in any retail store before it opens for business on the day, and calculate the inventory ratio of any first commodity in any retail store. The calculation formula is: Among them, PI ij is the inventory ratio of the jth first product in the ith retail store, NI ij is the inventory quantity of the j-th first commodity in the i-th retail store, where i is a positive integer from 1 to I, and I is the total number of retail stores; Obtain the total sales volume of any first commodity in any retail store within the second preset time from the material preparation time, and calculate the sales ratio of any first commodity in any retail store. The calculation formula is: Among them, PS ij is the sales ratio of the jth first commodity in the i-th retail store, TS ij is the total sales volume of the j-th first commodity in the i-th retail store, where i is a positive integer from 1 to I, and I is the total number of retail stores; The allocation ratio of any first commodity in any retail store is calculated based on the inventory ratio and the sales ratio of any first commodity in any retail store, and the calculation formula is: Among them, D ij is the allocation ratio of the jth first commodity in the i-th retail store, and is the parameter weight coefficient.

2. A digital operation method based on SaaS application according to claim 1, characterized in that: In the step 1, sending the first measured value to the SaaS platform includes: Step 111: The terminal device divides the first measured value into M1 data blocks according to a preset rule, and assigns a unique data block identifier to each data block, where M1 is a positive integer greater than 1; Step 112: search the communication path table based on the destination address of the SaaS platform, randomly select M2 communication paths, and send the M1 data blocks to M2 cloud servers on the M2 communication paths respectively, where M2≤M1; Step 113: After the sending is completed, the terminal device sends data sending information to the first server, wherein the data sending information includes terminal device information, a data identifier of the first measured value, a data block identifier of each data block, communication path information corresponding to each data block, and SaaS platform information; Step 114: the SaaS platform starts to inquire the first server from the second preset time point whether there is any information sent to the SaaS platform, and if so, obtains the terminal device information, the data identifier of the first measured value, the data block identifier of each data block, and the communication path information corresponding to each data block; Step 115: The SaaS platform obtains M1 data blocks from M2 cloud servers based on the data block identifier of each data block and the communication path information corresponding to each data block, and then uses the preset rule to obtain the first measured value according to the M1 data blocks.

3. According to a digital operation method based on SaaS application according to claim 1, it is characterized in that: The step 1 includes: At a predetermined time on the previous day, obtain the predicted values ​​of all supply influencing factors of any commodity in any retail store at each preset time point on the next day, wherein the supply influencing factors at least include the sales influencing factors, the commodity number of any commodity, the actual sales volume of any commodity and the store logo of any retail store; Generate a predicted value time series for each supply influencing factor, and then based on the identifier of any supply influencing factor, input the predicted value time series of any supply influencing factor into a feature extraction module corresponding to any supply influencing factor to obtain a first feature vector of the predicted value time series of any supply influencing factor; All first feature vectors are input into the first prediction model, and the predicted delivery quantity of any commodity in any retail store at each supply time point is output.

4. A digital operation method based on SaaS application according to claim 3, characterized in that: The generation method of the first prediction model is: Step 121: randomly extract a second historical measured value sequence of any supply influencing factor of any commodity in one day from a historical distribution volume data table, wherein the historical distribution volume data table at least stores the actual supply volume of any commodity at each supply time point and the measured values ​​of all supply influencing factors at each preset time point; Step 122: based on the identifier of any supply influencing factor, input the second historical measured value sequence of any supply influencing factor into a feature extraction module corresponding to any supply influencing factor to obtain a second feature vector of the second historical measured value sequence of any supply influencing factor; Step 123: input all second eigenvectors into a neural network model for training to generate an initial prediction model; Step 124: Obtain a third historical measured value sequence different from the second historical measured value sequence from the historical delivery volume data table, generate a third feature vector according to the method of step 122, and then input all third feature vectors into the initial prediction model to obtain a first predicted delivery volume of any commodity at each supply time point; Step 125: calculate the deviation between the first predicted delivery quantity and the actual supply quantity at each delivery time point, adjust the parameters of the initial prediction model based on the deviation to generate a new initial prediction model, define the third historical measured value sequence as the second historical measured value sequence, and then repeat steps 124 and 125 until the deviation converges or reaches a preset number of training times, and use the new initial prediction model as the first prediction model.

5. A digital operation system based on SaaS application, used to implement the digital operation method based on SaaS application as described in any one of claims 1 to 4, characterized in that: include: A terminal device and a SaaS platform, wherein the SaaS platform includes a first judgment module, a first adjustment module, a second judgment module, and a second adjustment module; The terminal device is used to collect a first measured value of any of the retail store sales influencing factors at a preset time point, and send the first measured value to the SaaS platform; The first judgment module is used to judge whether the first measured value exceeds the predicted range, if not, the commodity is delivered according to the predicted delivery quantity, if it exceeds, the first adjustment module is entered; The first adjustment module is used to calculate the difference between the first measured value and the predicted range, and adjust the delivery quantity of all first commodities corresponding to the difference in any of the retail stores according to the difference, and obtain the adjusted delivery quantity of each first commodity respectively; The second judgment module is used to calculate the total adjusted delivery quantity of all retail stores for any first commodity, and judge whether the total adjusted delivery quantity exceeds the inventory quantity. If not, the commodity is delivered to any retail store according to the adjusted delivery quantity of the first commodity in any retail store; if it exceeds, the second adjustment module is entered; The second adjustment module is used to calculate the allocation ratio of any first commodity in any retail store, and distribute the commodity to any retail store based on the allocation ratio and the inventory quantity; The first adjustment module is further configured to perform the following steps before calculating the difference: Step 311: For any retail store, obtain second sales data of any commodity and a second measured value of the sales influencing factor at each preset time point in a first time period, wherein the first time period is a time period of a first preset time before the collection time of the first measured value; Step 312: Calculate the correlation between any commodity in any retail store and the sales influencing factor based on all second sales data and all second measured values ​​in the first time period; Step 313: when the absolute value of the correlation is greater than the first preset value and less than or equal to the second preset value, the influence is determined to be level one; when the absolute value of the correlation is greater than the second preset value and less than or equal to the third preset value, the influence is determined to be level two; when the absolute value of the correlation is greater than the third preset value, the influence is determined to be level three; Step 314: store the store ID of any retail store, the product ID of any product, the sales influencing factor, the relevance, and the influence in a first data table in a corresponding manner; The first adjustment module is further configured to perform the following steps: Obtaining the absolute value of the difference, and if the absolute value of the difference is less than or equal to a fourth preset value, traversing the first data table, and adjusting the delivery quantity of the commodity with the third level of influence; If the absolute value of the difference is greater than the fourth preset value and less than or equal to the fifth preset value, determine whether the delivery quantity of the commodity with the third degree of influence has been adjusted before the collection time; if not, traverse the first data table to adjust the delivery quantity of the commodities with the third and second degrees of influence; if yes, traverse the first data table to adjust the delivery quantity of the commodity with the second degree of influence; If the absolute value of the difference is greater than the fifth preset value, determining whether the delivery quantity of the commodities with the third and / or second degree of influence has been adjusted before the collection time, if not, traversing the first data table to adjust the delivery quantity of the commodities with the third, second and first degree of influence, if yes, traversing the first data table to adjust the delivery quantity of the commodities that have not been adjusted; The first adjustment module is further configured to perform the following steps: Step 321, obtaining a sequence of measured values ​​of the sales influencing factors at all preset time points on the current date of any retail store; Step 322: traverse the second data table based on the measured value sequence, and obtain the date of the first historical measured value sequence with the greatest similarity to the measured value sequence as a reference date, wherein the second data table is a table storing the historical measured values ​​of the sales influencing factors of any retail store, and the collection time of the first historical measured value sequence is the same as the collection time of the measured value sequence; Step 323: taking the historical delivery quantity of any retail store at the next delivery time point corresponding to any first commodity on the reference date as the adjusted delivery quantity of any first commodity of any retail store; In the second adjustment module, the calculation method of the distribution ratio is: Get the inventory of any first commodity in any retail store before it opens for business on the day, and calculate the inventory ratio of any first commodity in any retail store. The calculation formula is: Among them, PI ij is the inventory ratio of the jth first product in the ith retail store, NI ij is the inventory quantity of the j-th first commodity in the i-th retail store, where i is a positive integer from 1 to I, and I is the total number of retail stores; Obtain the total sales volume of any first commodity in any retail store within the second preset time from the material preparation time, and calculate the sales ratio of any first commodity in any retail store. The calculation formula is: Among them, PS ij is the sales ratio of the jth first commodity in the i-th retail store, TS ij is the total sales volume of the j-th first commodity in the i-th retail store, where i is a positive integer from 1 to I, and I is the total number of retail stores; The allocation ratio of any first commodity in any retail store is calculated based on the inventory ratio and the sales ratio of any first commodity in any retail store, and the calculation formula is: Among them, D ij is the allocation ratio of the jth first commodity in the i-th retail store, and is the parameter weight coefficient.

6. A computer storage medium, characterized in that: The computer storage medium stores program instructions, wherein when the program instructions are executed, the device where the computer storage medium is located is controlled to execute the digital operation method based on SaaS application according to any one of claims 1 to 4.

Images