To investigate surface movement and deformation characteristics due to continuous mining and continuous backfilling (CMCB)of coal under artificial lakes, laboratory and field coring mechanical tests were conducted on the CMCB area to verify the feasibility of the filling body. Based on the equivalent mining height probability integration method, the surface subsidence of the CMCB area was predicted. The height of the water-conducting fracture zone was analyzed using numerical simulation, and its results were compared with those of the probability integration method. The results show that the strength of the filling body is 5.063 MPa, which is higher than the designed strength of 2.0 MPa, ensuring safe mining.Owing to continuous mining and backfilling in the area, the maximum inclination value of the surface was 0.3 mm/m and the maximum horizontal deformation value of the surface was -0.2 mm/m, respectively, which is less than the range of grade Ⅰ damage to brick and concrete structures. The surrounding surface subsidence was gentle, and there was no safety hazard. The height of the water-conducting fracture zone was about 49.7 m, and the distance from the waterproof layer was about 160.3 m, indicating the safety of underwater coal mining. Results of the FLAC^3D numerical simulation and probability integration method were close, thereby verifying that the CMCB technology can effectively slow down surface movement and deformation.

The traditional maximum power point tracking (MPPT) algorithm is prone to fall into local optimization in the case of a multipeak photovoltaic array. The butterfly optimization algorithm has a strong global search capability and a relatively stable convergence process; however, it has not been widely used due to its low convergence accuracy. This paper proposes an MPPT algorithm that combines the improved butterfly optimization algorithm with the perturbation and observation method. The traditional butterfly optimization algorithm was optimized by introducing the chaotic mapping theory to improve the distribution of the initial butterfly population. Besides, the dynamic switching probability was used to optimize the switching strategy. Herein, first, the global search capability of the butterfly optimization algorithm was used to locate the range of the maximum power point, and then the small step size perturbation and disturbance observation method were used to accurately locate the maximum power point. This algorithm combines the advantages of the global optimization of the butterfly optimization algorithm and the precise optimization of the perturbation and observation method. Furthermore, Simulink simulation experiments were conducted, and the results were compared with the traditional butterfly optimization algorithm and particle swarm optimization algorithm. The results show that the improved algorithm can adapt to complex and changing light conditions and has certain advantages in both convergence accuracy and speed.