<p>This article explores software reliability growth models (SRGMs) incorporating testing effort, imperfect debugging, error generation, and change point. We acknowledge that fault detection rates can vary depending on the testing team’s effectiveness, program size, and software test ability. Therefore, we have considered two different fault detection rates while modeling SRGMs instead of taking a constant fault detection rate throughout the software development process. In reality, fault correction is tied to the rate at which faults are introduced. As a result, we have integrated two time-varying testing efforts, accounted for imperfect debugging, error generation, and introduced a change point into the proposed model. We utilized the least square estimation method to estimate the parameters of our proposed model, and the outcomes reveal its capability to offer reasonably accurate predictions. Furthermore, we explored the determination of release time, taking into account a cost function that changes over time during both the testing and operational phases.</p>