Software tests are fundamental in the reliability and quality of systems, contributing to their positioning in the market. Generating test data is a critical task, as exhaustive testing is costly in time and effort. An adequate design of the test cases, which contemplates a selection of adequate values, can detect a high number of defects. The effectiveness of the test cases is measured according to the number of errors they managed to detect. However, the proposals that address these issues with the use of heuristic algorithms focus on the reduction of generation time and different coverage criteria. This article presents a search-based optimization model for the generation of unit test suites that integrates different test case design techniques considering the significance of the values generated in the detection of errors. The significance of the paths is also taken into account, with the aim of obtaining test cases with greater potential to detect errors. The optimization model uses heuristic algorithms that maximize the coverage of the paths. The results of the experimentation are presented, which show that the proposal presented generates test suits with a high capacity to detect errors. For this, the effectiveness of the generated test suits to detect errors in the mutated code was evaluated.