This study addresses the issue of insufficient interdisciplinary adaptability in electrical engineering experimental teaching within the context of new engineering disciplines. To resolve the structural contradiction between the current standardized teaching content system and students' personalized development needs, a layered experimental teaching scheme based on a lithium-ion battery over-discharge protection system is designed.

The research is conducted within the engineering context of satellite power systems. The system comprises a voltage comparison module, a threshold comparison module, and a relay drive module. It is supported by a three-level experimental teaching system consisting of a “basic layer,” an “advanced layer,” and an “innovation layer,” forming a competency development pathway of “system cognition, professional deepening, and innovative breakthrough.” As a compulsory module for students from different majors, the basic experiment guides students to decompose system-level technical indicators (such as voltage detection range and over-discharge protection threshold) into achievable circuit module design parameters through problem decomposition and index analysis of engineering tasks, thereby strengthening theoretical cognition and basic practical ability. The advanced experiment and the innovative experiment implement differentiated designs according to professional characteristics. The advanced experiment introduces a “two-out-of-three” redundant architecture design based on the characteristics of device failure in extreme space environments. Through the reliability design of both the sampling end and the execution end, students' understanding of hardware reliability is strengthened. The innovative experiment addresses the limitations of the above scheme in performing energy early warning and mitigating relay contact adhesion. By integrating a multi-level energy management software algorithm, the system achieves collaborative optimization of the hardware protection circuit and the software-based intelligent algorithm, with emphasis on enhancing students' algorithm implementation capability in the field of embedded systems. Students with stronger learning capabilities can further carry out full-process engineering training. Through comprehensive practice from circuit design to system implementation, the final output conforms to the functional model required by engineering application standards.

Using Multisim software, the designs of the basic and advanced experiments were tested. (1) When the battery voltage is 28 V, the corresponding output is 5 V. The measured output characteristics exhibit a good linear relationship within the range of 21–29.4 V, 4 V, meeting the voltage detection range requirement of Task 1. (2) When the battery voltage exceeds 21.0117 V, the output is low (relay closed), and when the battery voltage falls below 20.6322 V, the output is high (relay activated), satisfying the threshold level requirements in Task 2. (3) When any sampling channel device fails, the system voting output can still maintain the correct voltage value based on the remaining two normal sampling channels, meeting the Task 3 requirement of allowing one sampling failure. (4) When any drive circuit device fails, the transistor array can still ensure correct relay operation, thereby meeting the Task 3 requirement of allowing one drive failure. (5) The software algorithm in the innovative experiment was verified using a software parameter injection testing method. The results show that when the battery voltage drops to the first warning value, the system correctly triggers the energy warning and disconnects the high-power load. If the voltage continues to decrease to the second warning value, the system further disconnects non-essential loads and finally performs power-off protection, satisfying the multi-level voltage protection requirements specified in Task 4.

As a comprehensive practical component of the electrical engineering experimental course, the satellite power over-discharge protection system designed in this study is arranged at the end of the course with a total of 6 class hours. Students are required to integrate their prior knowledge of circuit analysis, instrument operation, Arduino/C programming, and related subjects to complete full-process training from theory to engineering implementation. Guided by engineering problems, the system establishes multi-level and extensible experimental tasks, achieving the organic integration of students' independent choice and engineering capability development. It also provides personalized development pathways for students from different disciplinary backgrounds.