The panel performance of a prefabricated cabin-type substation under the impact of fires plays a vital role in the normal operation of the substation. However, current evaluations of the panel performance of substations under fire still focus on fire resistance tests, which seldom consider the relationship between fire behavior and the mechanical load of the panel under the impact of fires. Aiming at the complex and uncertain relationship between the thermal and mechanical performance of the substation panel under impact of fires, this paper proposes a machine learning method based on a BP neural network. First, the fire resistance test and the stress test of the panel is carried out, then a machine learning model is established based on the BP neural network. According to the collected data, the model parameters are obtained through a series of training and verification processes. Meanwhile, the correlation between the panel performance and fire resistance was obtained. Finally, related parameters are input into the thermal–mechanical coupling evaluation model for the substation panel performance to evaluate the fire resistance performance of the substation panel. To verify the correctness of the established model, numerical simulation of the fire test and stress test of the panel is conducted, and numerical simulation samples are predicted by the trained model. The results show that the prediction curve of neural network is closer to the real results compared with the numerical simulation, and the established model can accurately evaluate the thermal–mechanical coupling performance of the substation panel under fire.

Joint connections are critical to the overall performance of prefabricated structures. This paper proposes a novel reverse-rotation locking sleeve connection method, designed to ensure the safety of joint engineering while optimizing construction processes, improving operational efficiency, and endowing the joints with excellent seismic energy dissipation performance. To evaluate the performance of this connection method, quasi-static tests under displacement-controlled lateral loading were designed and conducted on three reinforced concrete column specimens (Specimen A: conventional reinforcement–cast-in-place monolithic; Specimen B: conventional reinforcement–reverse-rotation locking sleeve connected; Specimen C: enhanced reinforcement–reverse-rotation locking sleeve connected). The failure modes, hysteretic characteristics, skeleton curves, ductility, energy dissipation capacity, load-bearing capacity, and stiffness degradation patterns of the specimens were systematically examined. The results indicate that Specimen B exhibited the most severe damage extent, while Specimen A demonstrated the best integrity; in contrast, Specimen B showed significant and rapid degradation in energy dissipation capacity during the intermediate-to-late stages of testing; the hysteretic curves of Specimens B and C were full in shape, without obvious yield plateaus; the skeleton curves of all specimens exhibited S-shaped characteristics, and the peak loads of Specimens A and C corresponded to a lateral displacement of 21 mm, while that of Specimen B corresponded to a lateral displacement of 28 mm; compared to the cast-in-place monolithic Specimen A, the reverse-rotation locking sleeve–connected Specimens B and C showed increases in ultimate load under positive cyclic loading by 18.7% and 5.5%, respectively, and under negative cyclic loading by 40.8% and 2.0%, respectively; the ductility coefficients of all three specimens met the code requirement, being greater than 3.0 (Specimen A: 5.13; Specimen B: 3.56; Specimen C: 5.66), with Specimen C exhibiting a 10.3% improvement over Specimen A, indicating that the reverse-rotation locking sleeve–connected specimens possess favorable ductile performance; analysis revealed that the equivalent viscous damping coefficient of Specimen C was approximately 0.06 higher than that of Specimen A, meaning Specimen C had superior energy dissipation capacity compared to Specimen A, confirming that the reverse-rotation locking sleeve connection can effectively absorb seismic energy and enhance the seismic and energy dissipation characteristics of the specimens. The load-bearing capacity degradation coefficients of all specimens fluctuated between 0.83 and 1.01, showing an initial stable phase followed by a gradual declining trend; the stiffness degradation coefficients exhibited rapid initial decline, followed by a deceleration in the attenuation rate, and eventual stabilization. This indicates that the reverse-rotation locking sleeve-connected specimens can maintain relatively stable strength levels and favorable seismic performance during the plastic deformation stage.