CN-121070749-B - System interface verification method and system based on artificial intelligence

CN121070749BCN 121070749 BCN121070749 BCN 121070749BCN-121070749-B

Abstract

The invention provides a system interface verification method and system based on artificial intelligence, and relates to the technical field of interface verification, wherein the method comprises the steps of obtaining historical concurrent user numbers, historical system performance data and historical load influence data; the method comprises the steps of obtaining system interface architecture data and historical system interface architecture data, obtaining real-time load influence data, determining a concurrent user number relation function according to the historical concurrent user number and the historical load influence data, determining a predicted concurrent user number of a prediction period according to the real-time load influence data and the concurrent user number relation function, obtaining a trained system performance prediction model, processing the predicted concurrent user number and the system interface architecture data according to the trained system performance prediction model to obtain a predicted system performance coefficient, and generating a system interface verification report according to the predicted system performance coefficient. According to the invention, the accuracy of system interface verification can be improved.

Inventors

SUN FENG
CHEN YIHAO
QIU ZHENXING

Assignees

上海喜数信息科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251103

Claims (4)

1. A system interface verification method based on artificial intelligence, comprising: acquiring historical concurrent user numbers, historical system performance data and historical load influence data in a plurality of historical periods; Acquiring system interface architecture data and historical system interface architecture data; Acquiring real-time load influence data; determining a concurrency user number relation function according to the historical concurrency user number and the historical load influence data; determining a predicted concurrent user number of a prediction period according to the real-time load influence data and the concurrent user number relation function; Training a system performance prediction model according to the historical system interface architecture data, the historical concurrent user number and the historical system performance data to obtain a trained system performance prediction model; Processing the predicted concurrent user number and the system interface architecture data according to the trained system performance prediction model to obtain a predicted system performance coefficient; generating a system interface verification report according to the predicted system performance coefficient; Determining a concurrency user number relationship function according to the historical concurrency user number and the historical load influence data, wherein the determining comprises the following steps: acquiring the number of active reference users of the history month corresponding to each history period; according to the historical load influence data, determining historical advertisement putting quantity, historical weather temperature and historical weather rainfall; Determining a concurrent user number relationship function according to the historical concurrent user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall; Determining a concurrency user number relationship function according to the historical concurrency user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall, wherein the concurrency user number relationship function comprises the following steps of: determining a pending coefficient equation of the concurrent user number relationship function, wherein, For the number of history concurrency users for the i-th history period, For a preset number of concurrent users threshold value, The reference number of users is active for the history month corresponding to the i-th history period, Historical advertising volume for the ith historical period, To preset the advertisement delivery amount threshold value, For the i-th historical period of historical weather temperature, For a preset weather temperature threshold value, For the historical weather rainfall for the i-th historical period, For a preset weather rainfall threshold, 、、、、、、、 And A first coefficient to be determined for the coefficient to be determined equation; Solving the first coefficient to be determined according to the historical concurrent user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall to obtain a solution value of the first coefficient to be determined; determining a concurrent user number relation function according to the solution value of the first coefficient to be determined and the equation of the coefficient to be determined; Training a system performance prediction model according to the historical system interface architecture data, the historical concurrent user number and the historical system performance data to obtain a trained system performance prediction model, wherein the training comprises the following steps: Determining a historical system performance coefficient according to the historical system performance data; processing the historical concurrent user number and the system interface architecture data according to the historical system performance prediction model to obtain a sample system performance coefficient, wherein the historical system interface architecture data comprises a historical thread pool maximum thread number, a historical database connection pool maximum connection number and a historical containerized copy number; Determining a training loss function of a system performance prediction model according to the historical system performance coefficient, the sample system performance coefficient, the historical concurrent user number and the historical system interface architecture data; training the system performance prediction model according to a training loss function of the system performance prediction model to obtain a trained system performance prediction model; Determining a training loss function of a system performance prediction model according to the historical system performance coefficient, the sample system performance coefficient, the historical concurrent user number and the historical system interface architecture data, wherein the training loss function comprises the following steps of: determining training loss functions for a predictive model of system performance , wherein, The historical system coefficient of performance for the i-th historical period, The sample system coefficient of performance for the i-th history period, For the number of history concurrency users for the i-th history period, For a preset number of concurrent users threshold value, The maximum number of threads for the history thread pool for the i-th history period, For the preset thread number threshold value, The maximum number of connections to the pool for the history database for the i-th history period, For a preset maximum connection number threshold value, The number of copies is containerized for the history of the i-th history period, For a preset copy number threshold value, n is the number of history periods, i is less than or equal to n, and both i and n are positive integers.
2. The artificial intelligence based system interface verification method of claim 1, wherein determining a predicted number of concurrent users for a prediction period based on the real-time load impact data and the concurrent user relationship function comprises: Acquiring the current active reference user number corresponding to the prediction period; Determining a planned advertisement delivery amount according to the real-time load influence data; Determining weather forecast data according to the real-time load influence data; according to the weather forecast data, determining a predicted weather temperature and a predicted weather rainfall; And determining the predicted concurrent user number of the prediction period according to the relation function of the current active reference user number, the planned advertisement putting amount, the predicted weather temperature, the predicted weather rainfall and the concurrent user number.
3. The artificial intelligence based system interface verification method of claim 1, wherein determining historical system performance coefficients from the historical system performance data comprises: according to the historical system performance data, determining historical response time, historical application instance indexes, historical operating system CPU utilization rate and historical database connection number; Determining a performance coefficient of a historical user perception layer according to the historical response time; determining a historical application layer performance coefficient according to the historical application instance index; determining a performance coefficient of a historical system resource layer according to the CPU utilization rate of the historical operating system; Determining a performance coefficient of a historical downstream dependent layer according to the connection number of the historical database; and determining the historical system performance coefficient according to the historical user perception layer performance coefficient, the historical application layer performance coefficient, the historical system resource layer performance coefficient and the historical downstream dependent layer performance coefficient.
4. A system interface verification system based on artificial intelligence for performing the method of any one of claims 1-3, comprising: The historical data module is used for acquiring historical concurrent user numbers, historical system performance data and historical load influence data in a plurality of historical periods; the architecture data module is used for acquiring system interface architecture data and historical system interface architecture data; the influence data module is used for acquiring real-time load influence data; The relation function module is used for determining a relation function of the concurrent user number according to the historical concurrent user number and the historical load influence data; The prediction user module is used for determining the predicted concurrent user number of the prediction period according to the real-time load influence data and the concurrent user number relation function; the model training module is used for training the system performance prediction model according to the historical system interface architecture data, the historical concurrent user number and the historical system performance data to obtain a trained system performance prediction model; The prediction coefficient module is used for processing the predicted concurrent user number and the system interface architecture data according to the trained system performance prediction model to obtain a prediction system performance coefficient; and the verification report module is used for generating a system interface verification report according to the predicted system performance coefficient.

Description

System interface verification method and system based on artificial intelligence Technical Field The present invention relates to the field of interface verification technologies, and in particular, to a system interface verification method and system based on artificial intelligence. Background In the related art, system interface verification may be performed based on artificial intelligence, however, it is difficult for the related art to predict the load of the system interface and perform predictive performance test based on the predicted load, i.e., it is difficult to predict the interface performance bottleneck situation at future loads. The information disclosed in the background section of the application is only for enhancement of understanding of the general background of the application and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art. Disclosure of Invention The invention provides a system interface verification method and system based on artificial intelligence, which can solve the technical problem that the related technology is difficult to predict the interface performance bottleneck situation under the future load. According to a first aspect of the present invention, there is provided an artificial intelligence based system interface verification method, comprising: acquiring historical concurrent user numbers, historical system performance data and historical load influence data in a plurality of historical periods; Acquiring system interface architecture data and historical system interface architecture data; Acquiring real-time load influence data; determining a concurrency user number relation function according to the historical concurrency user number and the historical load influence data; determining a predicted concurrent user number of a prediction period according to the real-time load influence data and the concurrent user number relation function; Training a system performance prediction model according to the historical system interface architecture data, the historical concurrent user number and the historical system performance data to obtain a trained system performance prediction model; Processing the predicted concurrent user number and the system interface architecture data according to the trained system performance prediction model to obtain a predicted system performance coefficient; and generating a system interface verification report according to the predicted system performance coefficient. According to the invention, a concurrency user number relation function is determined according to the historical concurrency user number and the historical load influence data, and the method comprises the following steps: acquiring the number of active reference users of the history month corresponding to each history period; according to the historical load influence data, determining historical advertisement putting quantity, historical weather temperature and historical weather rainfall; And determining a concurrent user number relation function according to the historical concurrent user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall. According to the invention, a concurrency user number relation function is determined according to the historical concurrency user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall, and the concurrency user number relation function is determined according to the formula: , determining a pending coefficient equation of the concurrent user number relationship function, wherein, For the number of history concurrency users for the i-th history period,For a preset number of concurrent users threshold value,The reference number of users is active for the history month corresponding to the i-th history period,Historical advertising volume for the ith historical period,To preset the advertisement delivery amount threshold value,For the i-th historical period of historical weather temperature,For a preset weather temperature threshold value,For the historical weather rainfall for the i-th historical period,For a preset weather rainfall threshold,、、、、、、、AndA first coefficient to be determined for the coefficient to be determined equation; Solving the first coefficient to be determined according to the historical concurrent user number, the historical month active reference user number, the historical advertisement putting amount, the historical weather temperature and the historical weather rainfall to obtain a solution value of the first coefficient to be determined; and determining a concurrent user number relation function according to the solution value of the first coefficient to be deter