A method for prioritizing test cases of upper-level multi-objectives

Abstract

Aiming at the problem of prioritization of multi-objective test cases in regression testing, the invention discloses a method for prioritization of upper-level multi-objective test cases. In this method, the orderly sequence of test case numbers is used as the particle code, and the set of test case number sequences is used as the particle swarm to randomly generate the initial population; the average branch coverage and effective execution time of the program to be tested in the test case sequence are used as the fitness evaluation function ; Then use the epistasis crossover method to generate new individuals, and use the particles in the non-dominated solution set as the global optimal particles; finally, when the number of iterations reaches the maximum number of iterations, the individuals in the non-dominated solution set are the optimal multi-objective sorting results. Compared with the existing methods, the present invention provides a multi-objective test case prioritization method with a wide range of non-dominated solution sets and higher fitness values. This method helps to find software defects as early as possible in the regression test process, Reduce the cost of testing.

Description

technical field [0001] The invention belongs to the technical field of software testing, in particular to the technical field of software regression testing, and specifically relates to a prioritization method for upper-level multi-objective test cases. Background technique [0002] In the process of software evolution, Test Case Prioritization (TCP) technology, as an efficient and practical regression testing technology, achieves higher testing efficiency by sorting test cases according to certain test objectives, which is helpful for improving defect It is of great significance to improve the early detection rate and reduce the cost of testing. With the continuous improvement of industrial testing requirements, it is no longer enough to optimize the sequence of test cases for a single test target, because the actual testing process needs to consider the impact of various factors on software quality, such as testing cost, time and Code modification and other factors, Multi...

AI Technical Summary

Problems solved by technology

Deb Kalyanmoy et al. proposed a fast non-dominated sorting genetic algorithm (Non-dominated Sorting Genetic Algorithm II, NSGA-II) with an elite strategy to solve multi-objective optimization problems. Although NSGA-II runs fast and has a good convergence of the solution set, but Operations such as selection, crossover, and mutation in genetic algorithms are relatively complex, and the performance of the algorithm is not ideal

The particle swarm optimization (Multi-ObjectiveParticle Swarm Optimization, MOPSO) algorithm used to solve multi-objective optimization problems such as Tyagi Manika solves the prioritization of test cases, but the convergence of MOPSO is not good

Images