Dynamic calculation power adjustment method based on big data analysis

Abstract

The invention relates to a big data analysis-based computing power dynamic adjustment method. The method comprises the following steps of: intelligently analyzing the number of active users and the issuing number of orders corresponding to each household appliance sales platform in a current time segment by adopting an intelligent computing power demand analysis model corresponding to a computing power support mechanism; and based on the intelligent analysis result, respectively analyzing the default number of support servers in the current time segment for each household appliance sales platform according to the proportion again. According to the invention, for a computing power support scene in which computing power support is simultaneously provided by the same computing power support mechanism for each household appliance sales platform, a mode of combining static computing power distribution and dynamic computing power distribution can be adopted, so that reasonable flow of support computing power provided by the computing power support mechanism among the household appliance sales platforms is realized; therefore, under the background of limited computing power, the utilization rate of the supporting computing power is improved while the fact that each household appliance sales platform obtains the matching supporting computing power is ensured.

Description

Technical Field

[0001] The present invention relates to the field of big data analysis, and in particular to a method for dynamically adjusting computing power based on big data analysis. Background Art

[0002] With the widespread use and application of big data, big data analysis is gradually entering various sub-sectors. Big data analysis refers to the use of technical means to process and analyze massive, diverse, and rapidly generated data to unlock its potential value. Its significance lies in improving decision-making efficiency, discovering patterns, and optimizing resource allocation. Application areas include financial risk control, healthcare, smart cities, e-commerce, and intelligent manufacturing.

[0003] E-commerce is an important field for the application of big data analysis. For example, due to various reasons such as economic costs and management costs, general e-commerce sales platforms will not build their own server centers, that is, general e-commerce sales platforms will not provide computing power support themselves, but will rent servers from computing power support organizations to realize computing power support leasing. This makes it convenient to customize computing power support needs according to their own development dynamics. However, for various e-commerce sales platforms that are provided with computing power support by the same computing power support organization, because the computing power support organization is limited after all, while each e-commerce sales platform considers its own costs, the limited computing power of the computing power support organization and whether the computing power can be fully utilized become more important. Summary of the Invention

[0004] In order to solve the technical problems in the prior art, the present invention provides a method for dynamic adjustment of computing power based on big data analysis. For the computing power support scenario in which each e-commerce sales platform is provided with computing power support by the same computing power support organization, since the number of support servers provided by the computing power support organization is limited, that is, the supporting computing power that can be provided is limited, by adopting a combination of static computing power allocation and dynamic computing power allocation, the supporting computing power that can be provided by the computing power support organization can be reasonably flowed among the various e-commerce sales platforms. Specifically, the static computing power allocation is to respectively parse the default number of supporting servers for each e-commerce sales platform in proportion based on the number of registered users corresponding to each e-commerce sales platform. The dynamic computing power allocation is to re-parse the default number of supporting servers for each e-commerce sales platform in the current time segment in proportion based on the number of active users corresponding to each e-commerce sales platform in the current time segment. In this way, under the background of limited computing power, it is ensured that each e-commerce sales platform obtains matching supporting computing power while improving the utilization rate of supporting computing power.

[0005] According to the present invention, a method for dynamically adjusting computing power based on big data analysis is provided, the method comprising: Based on the number of registered users corresponding to each e-commerce sales platform, the default number of supported servers is resolved for each e-commerce sales platform in proportion. Each e-commerce sales platform is provided with computing power support by the same computing power support organization, and the computing power of each supporting server provided by the computing power support organization is the same; Obtain the number of active users and the number of orders placed for each e-commerce sales platform in each historical time segment before the current moment as the historical sales data corresponding to the e-commerce sales platform. Each historical time segment before the current moment and the current time segment with the current moment as the starting moment constitute a complete time interval, and the time interval length of each time segment is equal to the preset time interval length; Performing multiple training actions on the radial basis function neural network to obtain a radial basis function neural network after completing the multiple training actions and outputting the obtained radial basis function neural network as the intelligent computing power demand analysis model corresponding to the computing power support organization, wherein the number of training actions completed by the radial basis function neural network is monotonically positively correlated with the total number of all support servers owned by the computing power support organization; The intelligent computing power demand analysis model corresponding to the computing power support organization is used to intelligently analyze the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform; Based on the number of active users corresponding to each e-commerce sales platform in the current time segment, the default number of supported servers for each e-commerce sales platform in the current time segment is re-proportioned; Among them, based on the number of registered users corresponding to each e-commerce sales platform, the default number of supported servers for each e-commerce sales platform is respectively resolved in proportion. Each e-commerce sales platform is provided with computing power support by the same computing power support organization at the same time, including: accumulating the number of registered users corresponding to each e-commerce sales platform as the total number of registered users, and taking the ratio of the number of registered users corresponding to each e-commerce sales platform to the total number of registered users as the server number ratio corresponding to the e-commerce sales platform; Among them, based on the number of registered users corresponding to each e-commerce sales platform, the default number of supported servers for each e-commerce sales platform is analyzed proportionally. The computing power support provided to each e-commerce sales platform by the same computing power support organization also includes: multiplying the server number ratio corresponding to the e-commerce sales platform by the total number of all supported servers owned by the computing power support organization to obtain the default number of supported servers corresponding to the e-commerce sales platform.

[0006] It can be seen that the present invention has at least the following main inventive concepts: Inventive Concept A: For scenarios where each e-commerce sales platform is supported by the same computing power support organization, since the computing power support organization has a limited number of supporting servers and therefore limited computing power, a combination of static and dynamic computing power allocation is adopted to achieve a reasonable flow of computing power provided by the computing power support organization among the various e-commerce sales platforms. This ensures that each e-commerce sales platform obtains matching computing power within the context of limited computing power, while also improving the utilization rate of the supporting computing power. Inventive Concept B: Specifically, static computing power allocation involves re-proportionally calculating the default number of supporting servers for each e-commerce sales platform based on the number of registered users corresponding to each e-commerce sales platform. Dynamic computing power allocation involves re-proportionally calculating the default number of supporting servers for each e-commerce sales platform in the current time segment based on the number of active users corresponding to each e-commerce sales platform in the current time segment. Inventive Concept C: Using an intelligent computing power demand analysis model corresponding to a computing power support organization to synchronously and intelligently predict the number of active users corresponding to each e-commerce sales platform in the current time segment, wherein a radial basis function neural network is trained multiple times to obtain a radial basis function neural network after completing the multiple training actions and outputs the radial basis function neural network as the intelligent computing power demand analysis model corresponding to the computing power support organization, and the number of training actions completed by the radial basis function neural network is monotonically positively correlated with the total number of support servers owned by the computing power support organization; Inventive concept D: Introducing multiple basic information for synchronous intelligent prediction of the number of active users corresponding to each e-commerce sales platform in the current time segment. The multiple basic information include the length of a preset time interval, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform, thereby providing comprehensive and sufficient basic information for the above-mentioned synchronous intelligent prediction. DETAILED DESCRIPTION

[0007] The following is a detailed description of an embodiment of the method for dynamic adjustment of computing power based on big data analysis of the present invention.

[0008] <Primary embodiment of the present invention> The method for dynamically adjusting computing power based on big data analysis according to the primary embodiment of the present invention specifically includes the following steps: Based on the number of registered users corresponding to each e-commerce sales platform, the default number of supported servers is resolved for each e-commerce sales platform in proportion. Each e-commerce sales platform is provided with computing power support by the same computing power support organization, and the computing power of each supporting server provided by the computing power support organization is the same; For example, based on the number of registered users corresponding to each e-commerce sales platform, a default number of support servers is respectively resolved for each e-commerce sales platform in proportion, and each e-commerce sales platform is provided with computing power support by the same computing power support organization. The computing power of each support server provided by the computing power support organization is the same, including: each e-commerce sales platform is a different e-commerce sales platform in the same country or region, and each different e-commerce sales platform has its own customized e-commerce sales APP; Obtain the number of active users and the number of orders placed for each e-commerce sales platform in each historical time segment before the current moment as the historical sales data corresponding to the e-commerce sales platform. Each historical time segment before the current moment and the current time segment with the current moment as the starting moment constitute a complete time interval, and the time interval length of each time segment is equal to the preset time interval length; Specifically, each historical time segment before the current moment and the current time segment starting at the current moment form a complete time interval, and the time interval length of each time segment is equal to the preset time interval length, including: the value of the preset time interval length can be selected as 20 minutes; Performing multiple training actions on the radial basis function neural network to obtain a radial basis function neural network after completing the multiple training actions and outputting the obtained radial basis function neural network as the intelligent computing power demand analysis model corresponding to the computing power support organization, wherein the number of training actions completed by the radial basis function neural network is monotonically positively correlated with the total number of all support servers owned by the computing power support organization; The intelligent computing power demand analysis model corresponding to the computing power support organization is used to intelligently analyze the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform; Based on the number of active users corresponding to each e-commerce sales platform in the current time segment, the default number of supported servers for each e-commerce sales platform in the current time segment is re-proportioned; Among them, based on the number of registered users corresponding to each e-commerce sales platform, the default number of supported servers for each e-commerce sales platform is respectively resolved in proportion. Each e-commerce sales platform is provided with computing power support by the same computing power support organization at the same time, including: accumulating the number of registered users corresponding to each e-commerce sales platform as the total number of registered users, and taking the ratio of the number of registered users corresponding to each e-commerce sales platform to the total number of registered users as the server number ratio corresponding to the e-commerce sales platform; Wherein, based on the number of registered users corresponding to each e-commerce sales platform, respectively resolving the default number of supported servers for each e-commerce sales platform in proportion, and each e-commerce sales platform being provided with computing power support by the same computing power support organization at the same time further includes: multiplying the server number ratio corresponding to the e-commerce sales platform by the total number of all supported servers owned by the computing power support organization to obtain the default number of supported servers corresponding to the e-commerce sales platform; Among them, based on the number of active users corresponding to each e-commerce sales platform in the current time segment, re-proportionally analyzing the default number of supported servers for each e-commerce sales platform in the current time segment includes: accumulating the number of active users corresponding to each e-commerce sales platform in the current time segment as the aggregate number of active users, and taking the ratio of the number of active users corresponding to each e-commerce sales platform in the current time segment to the aggregate number of active users as the server number ratio of the e-commerce sales platform in the current time segment; And wherein, based on the number of active users corresponding to each e-commerce sales platform in the current time segment, re-analyzing the default number of supported servers for each e-commerce sales platform in the current time segment in proportion also includes: multiplying the server number ratio of the e-commerce sales platform in the current time segment by the total number of all supported servers owned by the computing power support organization to obtain the default number of supported servers of the e-commerce sales platform in the current time segment.

[0009] <Secondary Embodiment of the Invention> Compared with the primary embodiment of the present invention, the method for dynamically adjusting computing power based on big data analysis according to the secondary embodiment of the present invention further includes the following steps: According to the default number of servers supported by each e-commerce sales platform in the current time segment, the computing power support organization is configured in advance for each e-commerce sales platform to provide the supported computing power for each e-commerce sales platform in the current time segment; Specifically, the computing power support organization is pre-configured for each e-commerce sales platform based on the default number of supported servers for the current time segment. The supported computing power provided by each e-commerce sales platform for the current time segment includes: the current time segment is from 2:00 p.m. to 2:20 p.m., and the current time is 2:00 p.m.

[0010] <The present invention is to be implemented again> Compared with the first embodiment of the present invention, the method for dynamic adjustment of computing power based on big data analysis according to the second embodiment of the present invention specifically further includes the following steps: The wireless receiving interface at the computing power support institution is used to wirelessly receive the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment, and the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment are synchronously displayed on the computing power support institution side; Specifically, a wireless receiving interface at the computing power support organization is used to wirelessly receive the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment, and a synchronous display of the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment is executed on the computing power support organization side, including: the wireless receiving interface is based on a frequency division duplex communication mechanism.

[0011] Next, the specific steps of the method for dynamic adjustment of computing power based on big data analysis of the present invention will be further explained.

[0012] In the method for dynamic adjustment of computing power based on big data analysis according to various embodiments of the present invention: Performing multiple training actions on a radial basis neural network to obtain a radial basis neural network after completing the multiple training actions and outputting the obtained radial basis neural network as an intelligent computing power demand analysis model corresponding to the computing power support mechanism, wherein the number of training actions completed by the radial basis neural network is monotonically positively correlated with the total number of all support servers owned by the computing power support mechanism comprises: using a frequency mapping formula to represent a numerical mapping relationship of the monotonically positive correlation between the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support mechanism; Specifically, an FPGA chip may be selected to implement testing and simulation of a data processing process that uses a frequency mapping formula to represent a numerical mapping relationship that is monotonically positively correlated between the number of training actions completed by the radial basis function neural network and the total number of all support servers owned by the computing power support organization. The number mapping formula used to express the numerical mapping relationship of the monotonically positive correlation between the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization includes: in the number mapping formula, the total number of all support servers owned by the computing power support organization is the input value of the number mapping formula; And wherein, the numerical mapping relationship that is monotonically positively associated with the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization is represented by a frequency mapping formula includes: in the frequency mapping formula, the number of training actions completed by the radial basis neural network that is monotonically positively associated with the total number of all support servers owned by the computing power support organization is the output value of the frequency mapping formula.

[0013] And in the method for dynamic adjustment of computing power based on big data analysis according to various embodiments of the present invention: The intelligent computing power demand analysis model corresponding to the computing power support agency is used to intelligently analyze the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform, including: synchronously inputting the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform into the intelligent computing power demand analysis model corresponding to the computing power support agency; Among them, the intelligent computing power demand analysis model corresponding to the computing power support organization is used to intelligently analyze the number of active users and the number of orders issued of each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform. It also includes: running the intelligent computing power demand analysis model corresponding to the computing power support organization to obtain the number of active users and the number of orders issued of each e-commerce sales platform in the current time segment output by the intelligent computing power demand analysis model corresponding to the computing power support organization.

[0014] In addition, in the method for dynamic adjustment of computing power based on big data analysis, the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform are synchronously input into the intelligent computing power demand analysis model corresponding to the computing power support agency, including: using programmable logic devices to realize the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform are synchronously input into the intelligent computing power demand analysis model corresponding to the computing power support agency.

[0015] The method for dynamic adjustment of computing power based on big data analysis of the present invention is used to address the technical problem in the prior art that the computing power support scenario in which each e-commerce sales platform is provided with computing power support by the same computing power support organization cannot take into account both the full utilization of computing power and the limited computing power. For the computing power support scenario in which each e-commerce sales platform is provided with computing power support by the same computing power support organization, a combination of static computing power allocation and dynamic computing power allocation is adopted to achieve the reasonable flow of supporting computing power that can be provided by the computing power support organization among each e-commerce sales platform. Specifically, the static computing power allocation is to respectively resolve the default number of supporting servers for each e-commerce sales platform in proportion to the number of registered users corresponding to each e-commerce sales platform, and the dynamic computing power allocation is to re-resolve the default number of supporting servers for each e-commerce sales platform in the current time segment in proportion to the number of active users corresponding to each e-commerce sales platform in the current time segment. This ensures that each e-commerce sales platform obtains matching supporting computing power under the background of limited computing power while improving the utilization rate of supporting computing power.

[0016] Because certain changes may be made without departing from the scope of the present invention, the entire contents of the above description should be interpreted as illustrative and not merely literal. It should be appreciated by those skilled in the art that the order of steps or structures contained in the above description may be changed without departing from the scope of the present invention, and the descriptions contained herein are merely individual examples of the various possible descriptions of the present invention.

Claims

1. A method for dynamic adjustment of computing power based on big data analysis, characterized in that: The method comprises: Based on the number of registered users corresponding to each e-commerce sales platform, the default number of support servers for each e-commerce sales platform is respectively parsed in proportion, and each e-commerce sales platform is provided with computing power support by the same computing power support organization at the same time, and the computing power of each support server provided by the computing power support organization is the same, including: accumulating the number of registered users corresponding to each e-commerce sales platform as the total number of registered users, and taking the ratio of the number of registered users corresponding to each e-commerce sales platform to the total number of registered users as the server number ratio corresponding to the e-commerce sales platform; and multiplying the server number ratio corresponding to the e-commerce sales platform by the total number of all support servers owned by the computing power support organization to obtain the default number of support servers corresponding to the e-commerce sales platform; Obtain the number of active users and the number of orders placed for each e-commerce sales platform in each historical time segment before the current moment as the historical sales data corresponding to the e-commerce sales platform. Each historical time segment before the current moment and the current time segment with the current moment as the starting moment constitute a complete time interval, and the time interval length of each time segment is equal to the preset time interval length; Performing multiple training actions on the radial basis function neural network to obtain a radial basis function neural network after completing the multiple training actions and outputting the obtained radial basis function neural network as the intelligent computing power demand analysis model corresponding to the computing power support organization, wherein the number of training actions completed by the radial basis function neural network is monotonically positively correlated with the total number of all support servers owned by the computing power support organization; The intelligent computing power demand analysis model corresponding to the computing power support organization is used to intelligently analyze the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform; Based on the number of active users corresponding to each e-commerce sales platform in the current time segment, the default number of supported servers for each e-commerce sales platform in the current time segment is re-analyzed in proportion.

2. The method for dynamic adjustment of computing power based on big data analysis according to claim 1, characterized in that: Based on the number of active users corresponding to each e-commerce sales platform in the current time segment, the default number of supported servers for each e-commerce sales platform in the current time segment is re-proportioned, including: accumulating the number of active users corresponding to each e-commerce sales platform in the current time segment as the total number of active users, and taking the ratio of the number of active users corresponding to each e-commerce sales platform in the current time segment to the total number of active users as the server number ratio of the e-commerce sales platform in the current time segment; Among them, based on the number of active users corresponding to each e-commerce sales platform in the current time segment, re-analyzing the default number of supported servers for each e-commerce sales platform in the current time segment in proportion also includes: multiplying the server number ratio of the e-commerce sales platform in the current time segment by the total number of all supported servers owned by the computing power support organization to obtain the default number of supported servers of the e-commerce sales platform in the current time segment.

3. The method for dynamic adjustment of computing power based on big data analysis according to claim 2, characterized in that: The method further comprises: The computing power support organization is configured in advance for each e-commerce sales platform according to the default number of supported servers of each e-commerce sales platform in the current time segment to provide supported computing power for each e-commerce sales platform in the current time segment.

4. The method for dynamic adjustment of computing power based on big data analysis according to claim 2, characterized in that: The method further comprises: The wireless receiving interface at the computing power support organization is used to wirelessly receive the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment, and the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment are synchronously displayed on the computing power support organization side.

5. The method for dynamically adjusting computing power based on big data analysis according to any one of claims 2 to 4, characterized in that: Performing multiple training actions on the radial basis neural network to obtain the radial basis neural network after completing the multiple training actions and outputting it as the intelligent computing power demand analysis model corresponding to the computing power support organization, the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization are monotonically positively correlated, including: using a frequency mapping formula to represent the numerical mapping relationship of the monotonically positive correlation between the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization.

6. The method for dynamic adjustment of computing power based on big data analysis according to claim 5, characterized in that: The numerical mapping relationship that is monotonically positively correlated between the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization is represented by a frequency mapping formula, including: in the frequency mapping formula, the total number of all support servers owned by the computing power support organization is the input value of the frequency mapping formula.

7. The method for dynamic adjustment of computing power based on big data analysis according to claim 6, characterized in that: The use of a frequency mapping formula to express the numerical mapping relationship of the number of training actions completed by the radial basis neural network and the total number of all support servers owned by the computing power support organization, which is monotonically positively associated, includes: in the frequency mapping formula, the number of training actions completed by the radial basis neural network that is monotonically positively associated with the total number of all support servers owned by the computing power support organization is the output value of the frequency mapping formula.

8. The method for dynamically adjusting computing power based on big data analysis according to any one of claims 2 to 4, characterized in that: The intelligent computing power demand analysis model corresponding to the computing power support agency is used to intelligently analyze the number of active users and the number of orders issued by each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform. The method includes: synchronously inputting the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform into the intelligent computing power demand analysis model corresponding to the computing power support agency.

9. The method for dynamic adjustment of computing power based on big data analysis according to claim 8, characterized in that: The intelligent computing power demand analysis model corresponding to the computing power support organization is used to intelligently analyze the number of active users and the number of orders issued of each e-commerce sales platform in the current time segment based on the preset time interval length, the number of registered users corresponding to each e-commerce sales platform, and the historical sales data corresponding to each e-commerce sales platform. It also includes: running the intelligent computing power demand analysis model corresponding to the computing power support organization to obtain the number of active users and the number of orders issued of each e-commerce sales platform in the current time segment output by the intelligent computing power demand analysis model corresponding to the computing power support organization.