Research On Automatic Generation Of Path-oriented Test Data

The core of software automation testing is to generate test data efficiently and automatically.Man-made test data is a very time-consuming work,not only heavy workload,but also testing content is sightless.Automatic generation of test data can avoid the above defects,reduce the workload of the tester,and improve the efficiency of software testing significantly.In recent years,using intelligent optimization algorithms for automatic generation of test data of the implementation work has made great progress,the more popular is the study,test data generation based on genetic algorithm and genetic algorithm as initial encoding for the individual,the algorithm reuse is not strong.And the encoding and decoding takes up a large amount of CPU time,and the operation efficiency will be relatively low.It is found that the particle swarm optimization(PSO)algorithm is simple and efficient,and it is a promising algorithm in the field of automatic test data generation.Based on the research of test data generation technology and intelligent optimization algorithm,two improved optimization algorithms based on particle swarm optimization are proposed.The specific research work and results are as follows:1.A simplified improved particle swarm optimization(PSO)algorithm is proposed.The standard particle swarm optimization algorithm is used to reduce the order,simplify the evolution process of particles,and verify the feasibility of the optimization strategy through comparative experiments with standard particle swarm optimization.After improved particle swarm optimization,uncertain parameters only have inertia weight.The method of inertia weight selection is studied and analyzed.It is proposed that when the inertia weight is selected randomly in the proper domain,and the algorithm has the highest search ergodicity.2.An improved PSO-ACO algorithm is proposed to generate test data.In order to improve the problem that particle swarm algorithm is easy to fall into local optimum when solving problems,the optimized particle swarm optimization algorithm is combined with ant colony algorithm to condense the advantages of the two algorithms fully.This method avoids the defect that the particle swarm optimization is easy to fall into the local optimal,and is more simple and easy to understand.The experimental results show that the combined algorithm is used for automatic generation of test cases,and the advantages of the two algorithms are reflected to the greatest extent.Using feedback information to enhance global search capability,the number of individuals with high fitness for the whole solving process is not too large for the number of individuals with low fitness,and effectively improves the stability and balance of automatic test data generation,then it is easy to solve the "precocious" problem in the search process.3.A method of testing data generation based on K-Means and PSO combinational algorithm is proposed.In order to solve the problem that the particle swarm optimization is affected by the particle swarm size,the K-Means algorithm is introduced into the particle swarm optimization algorithm.During each iteration,the K-Means algorithm is used to divide the particle swarm to reduce the iterative number of the algorithm and improve the efficiency of the algorithm.The experimental results show:while applying the KPSO algorithm to the automatic generation of test data,it not only ensures the quality of the generated data,but also improves the efficiency of the algorithm,and saving the running time effectively makes the automation of test data more reliable,more reasonable and more practical.