Software Reliability Prediction Model Based on Improved Resampling Algorithm

Reliable software can provide security in finance, aerospace and other fields, and it can help find and repair potential software defects and problems, reducing the cost of software maintenance and support. The difficulty of software reliability is that modern software systems are becoming more and more complex, which makes it more difficult to predict and analyze software errors. In this paper, a software reliability evaluation model based on an improved particle filter algorithm is proposed. In the resamping stage, the selection probability of unique position particles is improved through the adaptive distance comparison step. In addition, the method further proposes the strategy of generating new particles, re-selecting low-weight particles and adding random noise, and finally obtaining small batches of new particles. The software reliability model and improved particle filter algorithm are combined to predict the number of software faults, and more accurate prediction results are obtained.