Demand-oriented code pre-configuration method, system and equipment and storage medium

Abstract

The invention relates to the technical field of code pre-configuration, in particular to a demand-oriented code pre-configuration method, system and equipment and a storage medium. Matching a plurality of performance function number sequences from the code function library according to the function requirements of the target scene; performing function demand sample frequent mining on the target scene to obtain a function demand input parameter and a function demand output identifier set; inputting the function demand input parameters into a plurality of function number sequences to obtain a plurality of function demand prediction outputs; traversing the plurality of function demand prediction outputs for outlier analysis to obtain a plurality of output prediction outlier factors; traversing the plurality of performance function number sequences to carry out calculation power demand prediction to obtain a plurality of expected calculation power; and based on the plurality of expected computing power and the plurality of output prediction outlier factors, sorting the plurality of performance function number sequences to obtain a target performance function number sequence, and executing code pre-configuration. The code development efficiency is improved, and the code practicability is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a demand-oriented code preconfiguration method, system, device and storage medium. Background Art

[0002] Current low-code development platforms primarily operate by pre-configuring individual functions and then combining them on a visual interface to complete development tasks. This approach only considers functional applicability, but ignores the adaptability of individual functions to the specific scenarios they address. This can lead to functions failing to meet specific scenario requirements, and the inefficient combination of functional modules. This results in a coarse-grained management system for low-code development platforms, resulting in low development efficiency and limited practicality. Summary of the Invention

[0003] The present invention aims to solve the technical problems of low code development efficiency and poor practicality in the prior art by providing a demand-oriented code preconfiguration method, system, device and storage medium.

[0004] The technical solution of the present invention to solve the above technical problems is as follows: In the first aspect, the present invention provides a demand-oriented code preconfiguration method, comprising: matching a number of function function number sequences from a code function library according to the functional requirements of a target scenario; performing frequency mining of function requirement samples on the target scenario to obtain function requirement input parameters and function requirement output identification sets; inputting the function requirement input parameters into the number of function function number sequences to obtain a number of function requirement prediction outputs; traversing the number of function requirement prediction outputs to perform outlier analysis according to the function requirement output identification set to obtain a number of output prediction outlier factors; traversing the number of function function number sequences to perform computing power demand prediction to obtain a number of expected computing powers; sorting the number of function function number sequences based on the number of expected computing powers and the number of output prediction outlier factors to obtain a target function function number sequence, and performing code preconfiguration.

[0005] Optionally, the code function library construction step includes: uploading a function function set from the management end, executing the numbering, and obtaining a function function number set, wherein any function of the function function set includes a function identifier; clustering the function function number set according to the function identifier to obtain several function identifier function function number sets; associating the several function identifier function function number sets with several function identifiers to generate the code function library.

[0006] Optionally, frequency mining of functional requirement samples is performed on the target scenario to obtain functional requirement input parameters and functional requirement output identification sets, including: collecting the functional requirement sample set of the target scenario, wherein any functional requirement sample in the functional requirement sample set includes functional requirement input monitoring data and functional requirement output monitoring data; performing similarity analysis on the functional requirement input monitoring data of the functional requirement sample set to obtain an input monitoring data similarity set; clustering the functional requirement sample set based on the input monitoring data similarity set and in combination with a similarity consistency threshold to obtain a multi-cluster functional requirement sample set; randomly extracting one functional requirement input monitoring data from each of the multi-cluster functional requirement sample sets to obtain multiple functional requirement input parameters, and adding them to the functional requirement input parameters; extracting multiple clusters of functional requirement output monitoring data from each of the multi-cluster functional requirement sample sets, performing central trend analysis, obtaining multiple functional requirement output identification sets, and adding them to the functional requirement output identification set, wherein the multiple functional requirement input parameters correspond one-to-one to the multiple functional requirement output identification sets.

[0007] Optionally, according to the function requirement output identification set, traverse the several function requirement prediction outputs to perform outlier analysis and obtain several output prediction outlier factors, including: extracting the first function requirement output identification set from the function requirement output identification set until the Qth function requirement output identification set, Q≥1, Q is an integer; extracting the first function requirement prediction output from the several function requirement prediction outputs, wherein the first function requirement prediction output includes the first function requirement sub-prediction output until the Qth function requirement sub-prediction output; analyzing the first output prediction outlier factor of the first function requirement sub-prediction output in the first function requirement output identification set; until analyzing the Qth output prediction outlier factor of the Qth function requirement sub-prediction output in the Qth function requirement output identification set; calculating the average of the first output prediction outlier factor and the Qth output prediction outlier factor, setting it as the first function function number sequence outlier factor, and adding it to the several output prediction outlier factors.

[0008] Among them, analyzing the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set includes: based on the first functional requirement sub-prediction output, performing similarity analysis with the first functional requirement output identification set to obtain an output data similarity set; based on the output data similarity set, analyzing the LOF outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set, and setting it as the first output prediction outlier factor.

[0009] Optionally, traversing the several function function number sequences to perform computing power demand prediction and obtain several expected computing powers, including: extracting a first function function number sequence from the several function function number sequences, wherein the first function function number sequence includes the first sequence function function to the Yth sequence function, Y≥1, and Y is an integer; traversing the first sequence function function to the Yth sequence function, extracting the first sequence historical separate call log to the Yth sequence historical separate call log, and performing centralized value evaluation of computing power loss record values ​​respectively to obtain the first sequence expected computing power to the Yth sequence expected computing power, and adding it to the first function function expected computing power; adding the first function function expected computing power to the several expected computing powers.

[0010] Optionally, based on the several expected computing powers and the several output prediction outlier factors, the several function function number sequences are sorted to obtain a target function function number sequence, including: configuring a first weight for the expected computing power and configuring a second weight for the output prediction outlier factor, wherein the first weight and the second weight are pre-configured by the management end, and the sum of the first weight and the second weight is equal to 1; de-dimensionalizing the several expected computing powers and the several output prediction outlier factors to obtain several de-dimensionalized values ​​of expected computing powers and several de-dimensionalized values ​​of output prediction outlier factors; weighting the several de-dimensionalized values ​​of expected computing powers and the several de-dimensionalized values ​​of output prediction outlier factors based on the first weight and the second weight to obtain several function function number adaptation coefficients; based on the several function function number adaptation coefficients, performing minimum value sorting on the several function function number sequences to obtain the target function function number sequence.

[0011] In a second aspect, the present invention provides a demand-oriented code preconfiguration system, comprising: Function matching module, used to match several function number sequences from the code function library according to the functional requirements of the target scenario; A function requirement sample mining module is used to perform function requirement sample frequency mining on the target scenario to obtain function requirement input parameters and function requirement output identification sets; A function requirement prediction module, configured to input the function requirement input parameters and the plurality of function function number sequences to obtain a plurality of function requirement prediction outputs; A prediction result outlier analysis module is used to traverse the plurality of function requirement prediction outputs according to the function requirement output identifier set to perform outlier analysis and obtain a plurality of output prediction outlier factors; A computing power demand prediction module is used to traverse the plurality of function number sequences to perform computing power demand prediction and obtain a plurality of expected computing powers; The function sorting module also sorts the function number sequences based on the expected computing power and the output prediction outlier factors, obtains the target function number sequence, and executes code preconfiguration.

[0012] In a third aspect, the present application provides an electronic device, comprising: a memory for storing a first computer software program; The processor is configured to read and execute the first computer software program, thereby implementing the demand-oriented code preconfiguration method in the first aspect.

[0013] In a fourth aspect, the present application provides a storage medium storing a second computer software program, which, when executed by a processor, implements the demand-oriented code preconfiguration method in the first aspect.

[0014] By implementing the present invention, it is possible to match a number of function number sequences from a code function library according to the functional requirements of a target scenario. By utilizing the characteristic of clustering by function identifier in the code function library, it is possible to quickly locate the function functions related to the functional requirements of the target scenario, thereby reducing the time of blindly searching for function functions, improving the starting point efficiency of development, and at the same time ensuring the basic functional relevance of the matched function functions to the requirements. By implementing the present invention, it is possible to perform frequency mining of functional requirement samples for the target scenario, obtain functional requirement input parameters and functional requirement output identification sets, gain an in-depth understanding of the functional requirement characteristics of the target scenario, clarify key input and output information, and provide an accurate basis for the verification and screening of subsequent functional functions, making subsequent steps more targeted and avoiding the problem of inapplicable functional functions due to a lack of in-depth understanding of the requirements. By implementing the present invention, it is possible to input the functional requirement input parameters and the functional function number sequences, obtain a number of functional requirement prediction outputs, and simulate the operation of the functional function number sequences in the target scenario through the actual input and output process, providing specific data support for subsequent outlier analysis, making it easier to determine whether these functional function number sequences can meet the output requirements of the target scenario; By implementing the present invention, it is possible to perform outlier analysis on the predicted outputs of several functional requirements according to the functional requirement output identification set, obtain several output prediction outlier factors, and quantitatively evaluate the difference between the output results of each functional function number sequence and the expected output of the target scenario. The smaller the outlier factor, the more the output of the sequence meets the requirements, which provides an important indicator for subsequent sorting and helps to screen out functional function combinations with more accurate outputs. By implementing the present invention, it is possible to traverse the plurality of function number sequences to predict computing power requirements, obtain a plurality of expected computing powers, and understand the computing power consumption of each function number sequence in advance, thereby avoiding operational problems caused by insufficient computing power in actual applications. At the same time, it also provides an indicator of resource consumption for subsequent sorting, which helps to select a function function combination with reasonable computing power requirements. By implementing the present invention, it is possible to sort the functional function number sequences based on the expected computing powers and the output prediction outlier factors, obtain the target functional function number sequence, execute code pre-configuration, and calculate the adaptation coefficient by weight and perform sorting, so as to select a functional function combination that meets the functional output requirements and has reasonable computing power, thereby improving the combination efficiency and adaptability of the functional modules, realizing more refined management, and laying a good foundation for subsequent development work.

[0015] To sum up, by implementing the present invention, it is possible to accurately screen out functional function combinations with high adaptability and reasonable computing power according to the specific needs of the target scenario, thereby improving the combination efficiency of functional modules, enhancing the adaptability of functional functions to the demand domain scenarios, and making the management of the low-code development platform more detailed, thereby improving the overall performance and practicality of the low-code development platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the flow of the demand-oriented code preconfiguration method provided by the present invention; Figure 2 A schematic diagram of the structure of the demand-oriented code pre-configuration system provided by the present invention; Figure 3 A schematic structural diagram of an electronic device provided by the present invention; Figure 4 A schematic diagram of a storage medium provided by the present invention; Figure 5 A schematic diagram of a box plot provided by the present invention.

[0017] In the accompanying drawings, the components represented by the reference numerals are as follows: Function matching module 11, function requirement sample mining module 12, function requirement prediction module 13, prediction result outlier analysis module 14, computing power requirement prediction module 15, function function sorting module 16, electronic device 300, memory 310, processor 320, first computer program 311, storage medium 400, second computer program 410. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Therefore, features specified as "first" or "second" may explicitly or implicitly include one or more features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0020] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0021] Example 1, as Figure 1 As shown, an embodiment of the present invention provides a demand-oriented code preconfiguration method, including: S100: Matching a number of function number sequences from a code function library according to the functional requirements of the target scenario; S200: Performing frequency mining of functional requirement samples on the target scenario to obtain functional requirement input parameters and functional requirement output identification sets; S300: Input the function requirement parameters and the function function number sequences to obtain a number of function requirement prediction outputs; S400: According to the functional requirement output identifier set, traverse the plurality of functional requirement prediction outputs to perform outlier analysis to obtain a plurality of output prediction outlier factors; S500: traversing the plurality of function number sequences to perform computing power demand prediction and obtain a plurality of expected computing powers; S600: Based on the expected computing powers and the output prediction outlier factors, sort the functional number sequences, obtain a target functional number sequence, and perform code preconfiguration.

[0022] In step S100 of the embodiment of the present application, the code function library construction step includes: Upload a function set from the management terminal, execute the numbering, and obtain a function number set, wherein any function in the function set includes a function identifier; Clustering the function function number sets according to the function identifiers to obtain a plurality of function identifier function function number sets; The plurality of function identification function number sets are associated with the plurality of function identifications and stored to generate the code function library.

[0023] In the embodiment of this application, the core purpose of building a code function library is to provide efficient and accurate basic data support for step S100. By standardizing the management and classification of function functions, it is ensured that the function functions related to the functional requirements of the target scenario can be quickly located in the subsequent matching process. This solves the problems of insufficient adaptability of function functions to scenarios and low combination efficiency in traditional low-code platforms, and lays an orderly and searchable function resource foundation for the entire demand-oriented code pre-configuration process.

[0024] First, you need to upload a function set containing multiple functions from the management client. Each function has a unique function identifier that describes its core functionality. Each uploaded function is then numbered to form a function number set. This unique identification allows each function to be easily managed and called.

[0025] For example, the management client uploads two types of function sets: a power grid status data preprocessing function set and a circuit fault prediction function set. The power grid status data preprocessing function set includes: a data cleaning function (function identifier: "Power Grid - Data Cleaning"); a data normalization function (function identifier: "Power Grid - Data Normalization"); and an outlier filtering function (function identifier: "Power Grid - Outlier Filtering").

[0026] The circuit fault prediction function set includes: a short circuit fault prediction function, whose function identifier is: circuit-short circuit prediction; an overload fault prediction function, whose function identifier is: circuit-overload prediction; and a ground fault prediction function, whose function identifier is: circuit-ground prediction.

[0027] Each function is then uniquely numbered. For example, the data cleaning function is numbered F001, the data normalization function is numbered F002, and the outlier filtering function is numbered F003; the short-circuit fault prediction function is numbered F004, the overload fault prediction function is numbered F005, and the ground fault prediction function is numbered F006. The resulting function number set is {F001, F002, F003, F004, F005, F006}.

[0028] Then, based on the functional identifiers of the functional functions, the function number sets are clustered. Specifically, function numbers with similar functions are grouped together based on the relevance of their functional identifiers. For example, the function identifiers "Power Grid - Data Cleaning," "Power Grid - Data Standardization," and "Power Grid - Outlier Filtering" all belong to the "Power Grid State Data Preprocessing" category, so numbers F001, F002, and F003 are clustered into the "Power Grid State Data Preprocessing Function Number Set." The function identifiers "Circuit - Short Circuit Prediction," "Circuit - Overload Prediction," and "Circuit - Grounding Prediction" all belong to the "Circuit Fault Prediction" category, so numbers F004, F005, and F006 are clustered into the "Circuit Fault Prediction Function Number Set."

[0029] Then, the two clustered function identification function function number sets are associated with the corresponding function identifications and stored: for example, the "grid state data preprocessing" identification is associated with {F001, F002, F003}, and the "circuit fault prediction" identification is associated with {F004, F005, F006}.

[0030] After the above association relationship is stored, a code function library is formed. Later, when the functional requirement of the target scenario is "grid status data preprocessing", the number set {F001, F002, F003} can be directly matched from the function library, providing accurate resources for the function number sequence matching in step S100.

[0031] In step S200 of the embodiment of the present application, frequency mining of functional requirement samples is performed on the target scenario to obtain functional requirement input parameters and a functional requirement output identifier set, including: Collecting a functional requirement sample set of the target scenario, wherein any functional requirement sample in the functional requirement sample set includes functional requirement input monitoring data and functional requirement output monitoring data; Performing similarity analysis on the functional requirement input monitoring data of the functional requirement sample set to obtain an input monitoring data similarity set; Based on the similarity set of the input monitoring data and in combination with a similarity consistency threshold, clustering the functional requirement sample set to obtain a multi-cluster functional requirement sample set; Randomly extracting a piece of functional requirement input monitoring data from each of the plurality of functional requirement sample sets, obtaining a plurality of functional requirement input parameters, and adding the obtained data into the functional requirement input parameters; Extract multi-cluster functional requirement output monitoring data from the multi-cluster functional requirement sample set respectively, perform central trend analysis, obtain multiple functional requirement output identification sets, and add them into the functional requirement output identification set, wherein the multiple functional requirement input parameters correspond one-to-one to the multiple functional requirement output identification sets.

[0032] In step S200 of the embodiment of the present application, the frequency mining of functional requirement samples is performed on the target scenario. The core purpose is to accurately extract the typical functional requirement input parameters and corresponding output identification sets in the scenario, and provide a standardized and representative reference benchmark for the subsequent verification of the adaptability of the functional function numbering sequence. By mining the frequency characteristics of the samples, it is ensured that the extracted input parameters and output identifications can reflect the general needs of the target scenario, avoiding the functional function adaptation deviation caused by the particularity of the samples, thereby improving the accuracy of subsequent functional function sorting, and solving the problem of insufficient adaptability between functional functions and scenarios in traditional low-code platforms.

[0033] During implementation, the first step is to collect a sample set of functional requirements for the target scenario. For example, if the target scenario is "grid state data preprocessing," then the first step is to collect a sample set of functional requirements for that scenario. Each sample must include "functional requirement input monitoring data," such as real-time monitoring values ​​of grid voltage, current, and frequency, and "functional requirement output monitoring data," such as preprocessed standardized voltage values ​​and abnormal current marking results.

[0034] For example, sample 1 may be: input monitoring data: voltage 220V, current 10A; output monitoring data: standardized voltage 1.0, no abnormal mark; sample 2 may be: input monitoring data: voltage 218V, current 12A; output monitoring data: standardized voltage 0.99, no abnormal mark, etc.

[0035] Next, it is necessary to perform similarity analysis on the functional requirement input monitoring data of the functional requirement sample set to obtain an input monitoring data similarity set.

[0036] For example, similarity analysis is performed on all input monitoring data in the functional requirement sample set in the above example, such as the voltage and current values ​​of different samples, and the similarity between any two input monitoring data is calculated to finally form an input monitoring data similarity set.

[0037] Specifically, the method for calculating the similarity of monitoring data can be: In the formula, A1 and A2 are the same attribute values ​​of two sets of input monitoring data, such as the voltage value of sample 1 and the voltage value of sample 2, and ϵ is a small constant (such as 0.001) to avoid the denominator being zero.

[0038] Assume that the input monitoring data of sample 1 is voltage 220V, current 10A; the input monitoring data of sample 2 is voltage 218V, current 12A. Then calculate the similarity of voltage attributes Similarly, the similarity of current attributes can be calculated, such as 0.984. The similarities of multiple attributes are averaged to obtain the monitoring data similarity, such as 0.992. The monitoring data similarity of multiple sets of input monitoring data is calculated to obtain the monitoring data similarity set.

[0039] Then, it is necessary to cluster the functional requirement sample set based on the similarity set of the input monitoring data and the similarity consistency threshold to obtain a multi-cluster functional requirement sample set. Among them, the similarity consistency threshold can be dynamically configured according to the task type of the target scene. For example, the threshold range can be 0.7-0.9. If the standard requirements for similarity judgment are high, a higher threshold can be selected within the range, and vice versa. Assume that the similarity consistency threshold is selected as 0.8, that is, samples with input monitoring data similarity ≥ similarity consistency threshold are grouped into the same cluster to form a multi-cluster functional requirement sample set. For example, samples 1 and 2 are classified into cluster 1 because of their similarity of 0.992≥0.8; samples 3 and 4 are classified into cluster 2, and so on.

[0040] Next, randomly extract a piece of functional requirement input monitoring data from each of the multiple clusters of functional requirement sample sets. This data serves as the representative of the cluster, i.e., the functional requirement input parameter. Multiple functional requirement input parameters are obtained and added to the functional requirement input parameter. For example, extract the input monitoring data of sample 1 (220V, 10A) from cluster 1, and extract the monitoring data of sample 3 (180V, 5A) from cluster 2 to form multiple functional requirement input parameters.

[0041] Furthermore, it is necessary to extract multi-cluster functional requirement output monitoring data from the multi-cluster functional requirement sample set respectively, perform central trend analysis on the numerical values ​​in the output monitoring data, such as the above-mentioned standardized voltage value, to obtain multiple functional requirement output identification sets, and add them into the functional requirement output identification set, wherein the multiple functional requirement input parameters correspond one-to-one to the multiple functional requirement output identification sets.

[0042] Optionally, the above analysis steps can be presented using a box plot. Taking the output monitoring data of a cluster of functional requirement sample sets in the above-mentioned "grid state data preprocessing" scenario and the preprocessed standardized voltage value as an example, the specific analysis is as follows: First, the data range needs to be determined. Assume that the cluster function requirement output monitoring data sample set contains 10 output monitoring data, namely the aforementioned standardized voltage values: 0.98, 1.02, 0.99, 1.01, 1.00, 0.97, 1.03, 0.99, 1.00, and 1.01.

[0043] like Figure 5 As shown, we then need to draw a box plot and calculate its quartiles: lower quartile (Q1) 0.99, median (Q2) 1.00, upper quartile (Q3) 1.01; the box of the box plot extends from Q1 to Q3, and the middle line is the median; and there are no obvious outliers, that is, no data exceeds 1.5 times the interquartile range.

[0044] It can be seen from the box plot that the cluster functional requirement output monitoring data set is concentrated between 0.99-1.01, with a median of 1.00. Therefore, the cluster functional requirement output identification set can be defined as "standardized voltage value 1.00±0.01" as the expected output benchmark corresponding to this type of input parameters.

[0045] The box plot can intuitively reflect the concentrated range and typical values ​​of the output data, and quickly determine the representative output identification set.

[0046] In step S300 of the embodiment of the present application, it is necessary to input the functional requirements as parameters and input the plurality of functional function number sequences to obtain a plurality of functional requirements prediction outputs.

[0047] The core purpose of step S300 of the embodiment of the present application is to simulate the operating effect of the function number sequence in the target scenario, and to generate corresponding function demand prediction outputs by substituting the representative function requirement input parameters extracted in S200 into the several function function number sequences matched by S100, so as to provide actual operation data support for the subsequent evaluation of the adaptability of these function function number sequences to the target scenario. Through this process, the abstract function function combination can be converted into a specific output result, so as to be compared and analyzed with the function requirement output identification set obtained in S200, laying the foundation for screening out a function function combination with high adaptability.

[0048] For example, in the aforementioned "grid state data preprocessing" scenario, the first step is to obtain the required functional input parameters from step S200. For example, the input parameters for cluster 1 are voltage 220V and current 10A, and the input parameters for cluster 2 are voltage 180V and current 5A. Then, from step S100, several functional function number sequences related to this scenario are matched, for example, sequence 1: {F001 (data cleaning), F002 (data standardization)}, and sequence 2: {F001 (data cleaning), F003 (outlier filtering)}.

[0049] For each function number sequence, call the functions in the sequence one by one, entering the corresponding function requirement input parameters. For example, if the input parameters of cluster 1, 220V and 10A, are entered into sequence 1, the raw data will first be processed through F001 (data cleaning) and then converted to a different format through F002 (data standardization). Similarly, the same input parameters will be entered into sequence 2 to obtain another set of processing results.

[0050] After receiving the input parameters, each function number sequence will output the corresponding processing results, that is, the function demand prediction output.

[0051] For example, the predicted output of sequence 1 for cluster 1 input parameters is a normalized voltage of 1.0 with no abnormality mark; the predicted output of sequence 2 is a cleaned voltage of 220V with no abnormality mark.

[0052] Finally, several functional requirement prediction outputs are obtained, each of which corresponds to a functional function number sequence and a functional requirement input parameter.

[0053] In step S400 of the embodiment of the present application, based on the functional requirement output identifier set, the functional requirement prediction outputs are traversed to perform outlier analysis to obtain a number of output prediction outlier factors, including: Extracting the first functional requirement output identification set to the Qth functional requirement output identification set from the functional requirement output identification set, where Q≥1 and Q is an integer; Extracting a first functional requirement prediction output from the plurality of functional requirement prediction outputs, wherein the first functional requirement prediction output includes a first functional requirement sub-prediction output to a Qth functional requirement sub-prediction output; Analyze the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set; Until analyzing the Qth output prediction outlier factor of the Qth functional requirement sub-prediction output in the Qth functional requirement output identification set; Calculate the mean of the first output prediction outlier factor and the Qth output prediction outlier factor, set it as the first functional number sequence outlier factor, and add it to the plurality of output prediction outlier factors.

[0054] In step S400 of the embodiment of the present application, this step is the initial stage of outlier analysis. The core purpose is to establish a one-to-one correspondence between the function requirement prediction output and the expected output, that is, the function requirement output identification set, to provide a structured data basis for the subsequent calculation of the output prediction outlier factor. In addition, by decomposing the predicted output into sub-items corresponding to the output identification set, it is ensured that the outlier analysis can accurately locate the deviation of each dimension, avoid the analysis error caused by data structure mismatch, and thus provide a refined comparison basis for evaluating the output adaptability of the function function number sequence.

[0055] Taking the aforementioned "grid state data preprocessing" scenario as an example, we first need to extract the first functional requirement output identification set to the Qth functional requirement output identification set, that is, from the functional requirement output identification set obtained in S200, extract the first to Qth group of output identification sets by cluster, where Q is the number of clusters. For example: The first functional requirement output identification set corresponds to the input parameters of cluster 1: {standardized voltage 1.00±0.01, no abnormal mark}; the second functional requirement output identification set corresponds to the input parameters of cluster 2: {standardized voltage 0.82±0.02, no abnormal mark}; for example, here Q=2, that is, there are 2 groups of output identification sets, corresponding to 2 input parameter clusters respectively.

[0056] Then, it is necessary to extract the corresponding functional requirement prediction output and decompose it into sub-items, that is, from the multiple functional requirement prediction outputs obtained in S300, extract the prediction output of a certain functional function number sequence, that is, the first functional requirement prediction output, and decompose it into sub-prediction outputs according to the input parameter cluster. For example: The prediction outputs for the first functional function number sequence, such as {F001, F002}, include: the first functional requirement sub-prediction output corresponding to the processing results of cluster 1 input parameters, namely {normalized voltage 1.02, no anomaly flag}; and the second functional requirement sub-prediction output corresponding to the processing results of cluster 2 input parameters, namely {normalized voltage 0.85, no anomaly flag}. The number of sub-prediction outputs matches the number of output identifier sets, Q = 2, to ensure a one-to-one correspondence.

[0057] Through the above steps, the dimensional alignment of the output identification set and the predicted output sub-items is achieved, laying a data matching foundation for the subsequent calculation of the outlier factor of each sub-prediction output, ensuring that the outlier analysis can accurately reflect the output adaptability of each functional function number sequence in different input scenarios.

[0058] In step S400 of the embodiment of the present application, analyzing the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set includes: Based on the first functional requirement sub-prediction output, similarity analysis is performed with the first functional requirement output identification set to obtain an output data similarity set; Based on the output data similarity set, the LOF outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set is analyzed and set as the first output prediction outlier factor.

[0059] In the embodiment of the present application, the core purpose of this step is to quantitatively evaluate the degree of deviation between the first functional requirement sub-prediction output and the first functional requirement output identification set. By calculating the first output prediction outlier factor, it is determined whether the sub-prediction output meets the expected output pattern of the target scenario. This process provides a key indicator for the subsequent screening of the adapted functional function number sequence, ensuring the output accuracy of the final selected functional function combination in a specific scenario, and solving the problem of poor matching between functional output and scenario requirements in traditional low-code platforms.

[0060] Taking the "grid status data preprocessing" scenario in the above steps as an example, we first need to determine the analysis object. For example, the first functional requirement sub-prediction output, that is, the processing result of a certain functional function sequence on the input parameters of cluster 1 can be: {standardized voltage 1.02, no abnormal mark}; the first functional requirement output identification set, that is, the expected output corresponding to cluster 1 can be: {standardized voltage 1.00±0.01, no abnormal mark} Then, it is necessary to perform output data similarity analysis to obtain the output data similarity set. For example, for each attribute in the output, such as the standardized voltage value and the abnormal mark, the output data similarity is calculated separately. The output data similarity of the standardized voltage attribute can be calculated using the aforementioned monitoring data similarity formula. At this time, the first functional requirement sub-prediction output is A1=1.02, the typical value of the first functional requirement output identification set is A2=1.00, ϵ=0.001, and the formula calculates that the output data similarity of the standardized voltage attribute is 0.9998. For the abnormal mark attribute, the first functional requirement sub-prediction output is "no abnormal mark", which is consistent with the identification set, and the similarity is recorded as 1.0. Then the output data similarity set = {0.9998, 1.0} Next, it is necessary to analyze the LOF outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set based on the output data similarity set, and set it as the first output prediction outlier factor.

[0061] Among them, LOF (local outlier factor) is an algorithm that determines the degree of outliers by comparing the density of samples with that of neighboring samples. The closer the value is to 1, the closer it is to the normal mode, and the larger the value is, the higher the degree of outliers.

[0062] Assume that in the dataset corresponding to the first functional requirement output identification set, there are multiple historical normal output samples such as {1.00, no anomaly}, {0.99, no anomaly}, etc., and put the first functional requirement sub-prediction output (1.02, no anomaly) into this dataset: Calculate the density ratio of the first functional requirement sub-prediction output to the neighborhood samples to obtain the LOF value. Since 1.02 is within the range of 1.00 ± 0.01 in the identification set and the anomaly markers match, the LOF value is close to 1, such as 1.05, indicating a low degree of deviation. This LOF value can be set as the first output prediction outlier factor, which is used to measure the degree of outlier of the first functional requirement sub-prediction output.

[0063] Through the above steps, the matching degree between the sub-prediction output and the expected output is converted into a quantifiable outlier factor, which provides an accurate basis for the subsequent comprehensive evaluation of the adaptability of the functional function number sequence, ensuring that the output of the selected function combination in specific scenarios is more in line with actual needs.

[0064] Further, it is necessary to continue analyzing the output prediction outlier factor until the Qth output prediction outlier factor of the Qth functional requirement sub-prediction output in the Qth functional requirement output identifier set is analyzed; For example, the outlier factor of the first sub-prediction output calculated through the above steps is 1.05, the outlier factor of the second sub-prediction output is 1.12, and so on.

[0065] Then, it is necessary to calculate the mean of the first output prediction outlier factor and the Qth output prediction outlier factor, set it as the first functional number sequence outlier factor, and add it to the several output prediction outlier factors.

[0066] In the above example, the outlier factor of the first function number sequence = (first outlier factor + second outlier factor) / Q = (1.05 + 1.12) / 2 = 1.085. 1.085 is then used as the overall outlier factor for the sequence and added to the output predicted outlier factor set.

[0067] In step S500 of the embodiment of the present application, computing power demand prediction is performed by traversing the plurality of function number sequences to obtain a plurality of expected computing powers, including: Extracting a first function function number sequence from the plurality of function function number sequences, wherein the first function function number sequence includes a first sequence number function function to a Yth sequence number function function, where Y≥1 and Y is an integer; Traverse the first sequence number function to the Yth sequence number function, extract the historical individual call logs of the first sequence number to the Yth sequence number, perform centralized value evaluation of the computing power loss record values, obtain the expected computing power of the first sequence number to the Yth sequence number, and add it to the expected computing power of the first function function; The expected computing power of the first functional function is added to the several expected computing powers.

[0068] In an embodiment of the present application, the purpose of step S500 is to quantitatively evaluate the computing power consumption requirements of each function numbering sequence, and to predict its expected computing power in the target scenario by analyzing the historical call logs of each function in the function numbering sequence, so as to provide key indicators of resource consumption dimensions for comprehensive screening of function function combinations with high adaptability and reasonable computing power in the subsequent step S600. This process solves the problem that traditional low-code platforms only focus on functional adaptability and ignore computing power costs, resulting in low combination efficiency and waste of resources. Among them, the historical call log records the calling status of each function function in various scenarios, including all records of calls alone and calls in combination with other functions. The historical separate call log only records the calling status of a function function when it runs independently, that is, the log when it is executed alone without being combined with other functions.

[0069] The content of the historical call log includes the function's unique identifier, call timestamp, input parameter characteristics, and hash rate loss records. For example, a set of historical individual call logs may include the function's unique identifier F001; input parameter characteristics including the voltage data field; hash rate loss records of 0.4GB of memory consumption; and the execution result identified as successful.

[0070] Continuing with the two function number sequences in the "grid status data preprocessing" scenario in the above steps, namely sequence 1 {F001, F002} and sequence 2 {F001, F003}, as an example, it is first necessary to extract the function number sequence and composition, and select the first function number sequence, such as sequence 1: {F001, F002}, which contains Y=2 function functions, namely the first-numbered function function F001 (data cleaning function) and the second-numbered function function F002 (data normalization function).

[0071] Next, you need to obtain historical call logs and evaluate the expected computing power of a single function. You need to traverse each function in the function number sequence, extract its historical individual call logs, and record the computing power loss value for each call, such as CPU utilization, memory consumption, etc. For example, if the memory space occupied by the function during runtime is used as a measure of computing power consumption, the historical computing power loss record values ​​for F001 can be [0.5GB, 0.6GB, 0.55GB, 0.58GB], and the historical computing power loss record values ​​for F002 can be [0.8GB, 0.75GB, 0.82GB, 0.78GB].

[0072] Then, perform a central tendency assessment on the recorded values ​​of each function, such as calculating the median or mean. Here, we take the median as an example: the median of the expected computing power of the first sequence number (F001) is 0.565GB, and the median of the expected computing power of the second sequence number (F002) is 0.79GB.

[0073] Next, we need to calculate the expected computing power of the first function number sequence as a whole. This is done by adding up the expected computing power of each function in the first function number sequence to obtain the expected computing power of the first function: For example, the expected computing power of sequence 1 = 0.565GB + 0.79GB = 1.355GB, and then 1.355GB is added to the expected computing power.

[0074] Then, repeat the above steps for sequence 2 {F001, F003}. Assuming that the expected computing power of F003 (outlier filtering) is 1.2GB, the expected computing power of sequence 2 is 0.565GB + 1.2GB = 1.765GB, which is also added to the expected computing power.

[0075] In step S600 of the embodiment of the present application, based on the several expected computing powers and the several output prediction outlier factors, the several functional function number sequences are sorted to obtain a target functional function number sequence, including: Configuring a first weight for the expected computing power and configuring a second weight for the output predicted outlier factor, wherein the first weight and the second weight are pre-configured by the management end, and the sum of the first weight and the second weight is equal to 1; De-dimensionalizing the plurality of expected computing powers and the plurality of output prediction outlier factors to obtain a plurality of de-dimensionalized values ​​of the expected computing powers and a plurality of de-dimensionalized values ​​of the output prediction outlier factors; weighting the plurality of expected computing power dimensionless values ​​and the plurality of output prediction outlier factor dimensionless values ​​based on the first weight and the second weight to obtain a plurality of function number adaptation coefficients; Based on the adaptation coefficients of the plurality of functional function numbers, the plurality of functional function number sequences are sorted for minimum values ​​to obtain the target functional function number sequence.

[0076] The purpose of step S600 in this embodiment of the present application is to find the optimal balance between functional adaptability and computing power cost. Through a multi-dimensional weighted evaluation, the combination that best suits the target scenario is selected from the candidate function number sequence. By quantitatively calculating the function number adaptation coefficient, the problem of resource waste or insufficient adaptability caused by considering only a single indicator when selecting function combinations in traditional low-code platforms is solved.

[0077] Continuing with the two function number sequences in the "grid status data preprocessing" scenario in the above steps, namely sequence 1 and sequence 2, as an example, at this time, after the calculation in the above steps, the outlier factor of sequence 1: {F001, F002} is 1.085, and the expected computing power is 1.355GB; the outlier factor of sequence 2: {F001, F003} is 1.35, and the expected computing power is 1.765GB.

[0078] Next, you need to configure a first weight for the expected computing power and a second weight for the output predicted outlier factor. The default first weight on the management side is 0.3-0.5, and the second weight is 0.5-0.7, with the sum of the weights equaling 1. The specific weight values ​​can be configured based on actual needs during implementation. If the energy consumption of code pre-configuration is more important, the first weight can be appropriately increased. If the accuracy of code pre-configuration is more important, the second weight can be appropriately increased, for example, the first weight can be 0.4 and the second weight can be 0.6.

[0079] Then, it is necessary to perform dimensionless processing on the several expected computing powers and the several output prediction outlier factors to obtain several expected computing power dimensionless values ​​and several expected computing power dimensionless values. Since the output prediction outlier factor is dimensionless and smaller values ​​indicate a more consistent output, reverse normalization is used. That is, the smaller the original value, the closer it is to 1 after normalization. For example, the dedimensionalized output prediction outlier factor is 1 − (output prediction outlier factor value − minimum value) / (maximum value − minimum value). The minimum value is the minimum value of the output prediction outlier factor, such as 1.085, and the maximum value is the maximum value of the output prediction outlier factor, such as 1.35. Therefore, the dedimensionalized output prediction outlier factor for sequence 1 is 1 − (1.085 − 1.085) / (1.35 − 1.085) = 1 − 0 = 1. Using the same method, the dedimensionalized output prediction outlier factor for sequence 2 is 0. The above steps are merely used to illustrate how to obtain the dedimensionalized output prediction outlier factor.

[0080] For the expected computing power, smaller values ​​indicate less resource consumption. Therefore, the same reverse normalization process is used to obtain the dimensionless value of the expected computing power. The calculation formula for the dimensionless value of the expected computing power is the same as the dimensionless value of the output prediction outlier factor, and will not be repeated here.

[0081] Next, it is necessary to weight the plurality of expected computing power dimensionless values ​​and the plurality of output prediction outlier factor dimensionless values ​​based on the first weight and the second weight to obtain a plurality of function number adaptation coefficients; Assuming that the dimensionless value of the output prediction outlier factor obtained through the above steps is 0.6 and the dimensionless value of the expected computing power is 0.57, then using the first and second weights described above, the function number adaptation coefficient = first weight × dimensionless value of expected computing power + second weight × dimensionless value of output prediction outlier factor = 0.4 × 0.57 + 0.6 × 0.6 = 0.228 + 0.36 = 0.588. A larger adaptation coefficient indicates better performance. Several function number adaptation coefficients were calculated using the above method.

[0082] Finally, among the several function function number adaptation coefficients, the function function number sequence with the largest function function number adaptation coefficient is selected as the target function function number sequence. Based on the target function function number sequence, code preconfiguration is performed.

[0083] Example 2, as Figure 2 As shown, based on the same inventive concept as the demand-oriented code preconfiguration method provided in Example 1, an embodiment of the present invention further provides a demand-oriented code preconfiguration system, including: Function matching module 11, used to match a number of function number sequences from a code function library according to the functional requirements of the target scenario; A function requirement sample mining module 12 is used to perform function requirement sample frequency mining on the target scenario to obtain function requirement input parameters and function requirement output identification sets; The function requirement prediction module 13 is used to input the function requirement input parameters and the plurality of function function number sequences to obtain a plurality of function requirement prediction outputs; A prediction result outlier analysis module 14 is configured to perform outlier analysis on the plurality of function requirement prediction outputs according to the function requirement output identifier set, and obtain a plurality of output prediction outlier factors; The computing power demand prediction module 15 is used to traverse the plurality of function number sequences to perform computing power demand prediction and obtain a plurality of expected computing powers; The function sorting module 16 also sorts the function number sequences based on the expected computing power and the output prediction outlier factors, obtains the target function number sequence, and executes code preconfiguration.

[0084] Furthermore, the function matching module 11 includes the following execution steps: Upload a function set from the management terminal, execute the numbering, and obtain a function number set, wherein any function in the function set includes a function identifier; Clustering the function function number sets according to the function identifiers to obtain a plurality of function identifier function function number sets; The plurality of function identification function number sets are associated with the plurality of function identifications and stored to generate the code function library.

[0085] Furthermore, the functional requirement sample mining module 12 includes the following execution steps: Collecting a functional requirement sample set of the target scenario, wherein any functional requirement sample in the functional requirement sample set includes functional requirement input monitoring data and functional requirement output monitoring data; Performing similarity analysis on the functional requirement input monitoring data of the functional requirement sample set to obtain an input monitoring data similarity set; Based on the similarity set of the input monitoring data and in combination with a similarity consistency threshold, clustering the functional requirement sample set to obtain a multi-cluster functional requirement sample set; Randomly extracting a piece of functional requirement input monitoring data from each of the plurality of functional requirement sample sets, obtaining a plurality of functional requirement input parameters, and adding the obtained data into the functional requirement input parameters; Extract multi-cluster functional requirement output monitoring data from the multi-cluster functional requirement sample set respectively, perform central trend analysis, obtain multiple functional requirement output identification sets, and add them into the functional requirement output identification set, wherein the multiple functional requirement input parameters correspond one-to-one to the multiple functional requirement output identification sets.

[0086] Furthermore, the prediction result outlier analysis module 14 includes the following execution steps: Extracting the first functional requirement output identification set to the Qth functional requirement output identification set from the functional requirement output identification set, where Q≥1 and Q is an integer; Extracting a first functional requirement prediction output from the plurality of functional requirement prediction outputs, wherein the first functional requirement prediction output includes a first functional requirement sub-prediction output to a Qth functional requirement sub-prediction output; Analyze the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set; Until analyzing the Qth output prediction outlier factor of the Qth functional requirement sub-prediction output in the Qth functional requirement output identification set; Calculate the mean of the first output prediction outlier factor and the Qth output prediction outlier factor, set it as the first functional number sequence outlier factor, and add it to the plurality of output prediction outlier factors.

[0087] The step of analyzing the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set includes: Based on the first functional requirement sub-prediction output, similarity analysis is performed with the first functional requirement output identification set to obtain an output data similarity set; Based on the output data similarity set, the LOF outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set is analyzed and set as the first output prediction outlier factor.

[0088] Furthermore, the computing power demand prediction module 15 includes the following execution steps: Extracting a first function function number sequence from the plurality of function function number sequences, wherein the first function function number sequence includes a first sequence number function function to a Yth sequence number function function, where Y≥1 and Y is an integer; Traverse the first sequence number function to the Yth sequence number function, extract the historical individual call logs of the first sequence number to the Yth sequence number, perform centralized value evaluation of the computing power loss record values, obtain the expected computing power of the first sequence number to the Yth sequence number, and add it to the expected computing power of the first function function; The expected computing power of the first functional function is added to the several expected computing powers.

[0089] Furthermore, the function sorting module 16 includes the following execution steps: Configuring a first weight for the expected computing power and configuring a second weight for the output predicted outlier factor, wherein the first weight and the second weight are pre-configured by the management end, and the sum of the first weight and the second weight is equal to 1; De-dimensionalizing the plurality of expected computing powers and the plurality of output prediction outlier factors to obtain a plurality of de-dimensionalized values ​​of the expected computing powers and a plurality of de-dimensionalized values ​​of the output prediction outlier factors; weighting the plurality of expected computing power dimensionless values ​​and the plurality of output prediction outlier factor dimensionless values ​​based on the first weight and the second weight to obtain a plurality of function number adaptation coefficients; Based on the adaptation coefficients of the plurality of functional function numbers, the plurality of functional function number sequences are sorted for minimum values ​​to obtain the target functional function number sequence.

[0090] Example 3, as Figure 3 As shown, based on the same inventive concept as the demand-oriented code preconfiguration method provided in the first embodiment, the embodiment of the present invention further provides an electronic device 300, including: Memory 310, for storing a first computer software program 311; The processor 320 is configured to read and execute the first computer software program 311 , thereby implementing a demand-oriented code preconfiguration method according to the first embodiment.

[0091] Memory refers to the device in a computer used to temporarily store running programs and data, including but not limited to random access memory (RAM), read-only memory (ROM), virtual memory, etc.

[0092] A processor is a component in a computer that is responsible for executing program instructions, processing data, and controlling the computer's operations. It includes but is not limited to general-purpose processors (GPPUs), graphics processing units (GPUs), and embedded processors.

[0093] Example 4, as Figure 4 As shown, based on the same inventive concept as the demand-oriented code preconfiguration method provided in Example 1, an embodiment of the present invention further provides a storage medium 400. Exemplarily, the storage medium can be a non-transitory computer-readable storage medium, and a second computer software program 410 is stored in the storage medium. When the second computer software program 410 is executed by the processor, it implements the demand-oriented code preconfiguration method as in Example 1.

[0094] Non-transitory storage media refers to storage media that can persist data after power failure, including but not limited to SSDs (solid-state drives), HDDs (hard disk drives), and flash memory devices (USB flash drives, memory cards).

[0095] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0096] Those skilled in the art will appreciate that embodiments of the present invention may provide methods, systems, or computer program products. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0097] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0098] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0099] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0100] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0101] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. Combined with the demand-oriented code preconfiguration method, it is characterized by: include: According to the functional requirements of the target scenario, match several function number sequences from the code function library; Performing frequency mining of functional requirement samples on the target scenario to obtain a functional requirement input parameter and a functional requirement output identifier set; Input the functional requirement parameters and the functional function number sequences to obtain a plurality of functional requirement prediction outputs; According to the functional requirement output identification set, traversing the plurality of functional requirement prediction outputs to perform outlier analysis and obtain a plurality of output prediction outlier factors; Traversing the plurality of function number sequences to predict computing power demand and obtain a plurality of expected computing powers; Based on the several expected computing powers and the several output prediction outlier factors, the several functional function number sequences are sorted to obtain a target functional function number sequence, and code preconfiguration is performed.

2. The demand-oriented code preconfiguration method according to claim 1, wherein: The code function library construction steps include: Upload a function set from the management terminal, execute the numbering, and obtain a function number set, wherein any function in the function set includes a function identifier; Clustering the function function number sets according to the function identifiers to obtain a plurality of function identifier function function number sets; The plurality of function identification function number sets are associated with the plurality of function identifications and stored to generate the code function library.

3. The demand-oriented code preconfiguration method according to claim 1, wherein: Performing frequency mining of functional requirement samples on the target scenario to obtain functional requirement input parameters and functional requirement output identifier sets includes: Collecting a functional requirement sample set of the target scenario, wherein any functional requirement sample in the functional requirement sample set includes functional requirement input monitoring data and functional requirement output monitoring data; Performing similarity analysis on the functional requirement input monitoring data of the functional requirement sample set to obtain an input monitoring data similarity set; Based on the similarity set of the input monitoring data and in combination with a similarity consistency threshold, clustering the functional requirement sample set to obtain a multi-cluster functional requirement sample set; Randomly extracting a piece of functional requirement input monitoring data from each of the plurality of functional requirement sample sets, obtaining a plurality of functional requirement input parameters, and adding the obtained data into the functional requirement input parameters; Extract multi-cluster functional requirement output monitoring data from the multi-cluster functional requirement sample set respectively, perform central trend analysis, obtain multiple functional requirement output identification sets, and add them into the functional requirement output identification set, wherein the multiple functional requirement input parameters correspond one-to-one to the multiple functional requirement output identification sets.

4. The demand-oriented code preconfiguration method according to claim 1, wherein: According to the functional requirement output identification set, the plurality of functional requirement prediction outputs are traversed to perform outlier analysis to obtain a plurality of output prediction outlier factors, including: Extracting the first functional requirement output identification set to the Qth functional requirement output identification set from the functional requirement output identification set, where Q≥1 and Q is an integer; Extracting a first functional requirement prediction output from the plurality of functional requirement prediction outputs, wherein the first functional requirement prediction output includes a first functional requirement sub-prediction output to a Qth functional requirement sub-prediction output; Analyze the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set; Until analyzing the Qth output prediction outlier factor of the Qth functional requirement sub-prediction output in the Qth functional requirement output identification set; Calculate the mean of the first output prediction outlier factor and the Qth output prediction outlier factor, set it as the first functional number sequence outlier factor, and add it to the plurality of output prediction outlier factors.

5. The demand-oriented code preconfiguration method according to claim 4, wherein: Analyzing the first output prediction outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set includes: Based on the first functional requirement sub-prediction output, similarity analysis is performed with the first functional requirement output identification set to obtain an output data similarity set; Based on the output data similarity set, the LOF outlier factor of the first functional requirement sub-prediction output in the first functional requirement output identification set is analyzed and set as the first output prediction outlier factor.

6. The demand-oriented code preconfiguration method according to claim 1, wherein: Traversing the plurality of function number sequences to perform computing power demand prediction, and obtaining a plurality of expected computing powers, including: Extracting a first function function number sequence from the plurality of function function number sequences, wherein the first function function number sequence includes a first sequence number function function to a Yth sequence number function function, where Y≥1 and Y is an integer; Traverse the first sequence number function to the Yth sequence number function, extract the historical individual call logs of the first sequence number to the Yth sequence number, perform centralized value evaluation of the computing power loss record values, obtain the expected computing power of the first sequence number to the Yth sequence number, and add it to the expected computing power of the first function function; The expected computing power of the first functional function is added to the plurality of expected computing powers.

7. The demand-oriented code preconfiguration method according to claim 1, wherein: Sorting the plurality of functional function number sequences based on the plurality of expected computing powers and the plurality of output prediction outlier factors to obtain a target functional function number sequence includes: Configuring a first weight for the expected computing power and configuring a second weight for the output predicted outlier factor, wherein the first weight and the second weight are pre-configured by the management end, and the sum of the first weight and the second weight is equal to 1; De-dimensionalizing the plurality of expected computing powers and the plurality of output prediction outlier factors to obtain a plurality of de-dimensionalized values ​​of the expected computing powers and a plurality of de-dimensionalized values ​​of the output prediction outlier factors; weighting the plurality of expected computing power dimensionless values ​​and the plurality of output prediction outlier factor dimensionless values ​​based on the first weight and the second weight to obtain a plurality of function number adaptation coefficients; Based on the adaptation coefficients of the plurality of functional function numbers, the plurality of functional function number sequences are sorted for minimum values ​​to obtain the target functional function number sequence.

8. Combined with the demand-oriented code pre-configuration system, it is characterized by: include: Function matching module, used to match several function number sequences from the code function library according to the functional requirements of the target scenario; A function requirement sample mining module is used to perform function requirement sample frequency mining on the target scenario to obtain function requirement input parameters and function requirement output identification sets; A function requirement prediction module, configured to input the function requirement input parameters and the plurality of function function number sequences to obtain a plurality of function requirement prediction outputs; A prediction result outlier analysis module is used to traverse the plurality of function requirement prediction outputs according to the function requirement output identifier set to perform outlier analysis and obtain a plurality of output prediction outlier factors; A computing power demand prediction module is used to traverse the plurality of function number sequences to perform computing power demand prediction and obtain a plurality of expected computing powers; The function sorting module also sorts the function number sequences based on the expected computing power and the output prediction outlier factors, obtains the target function number sequence, and executes code preconfiguration.

9. An electronic device, characterized in that: include: a memory for storing a first computer software program; A processor is configured to read and execute the first computer software program, thereby implementing the demand-oriented code preconfiguration method according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores a second computer software program, which, when executed by the processor, implements the demand-oriented code preconfiguration method according to any one of claims 1 to 7.

Images