Abstract: ‍ ‍Backscatter communication technology is characterized by low cost and low power consumption， and it exhibits significant potential for widespread use in large-scale Internet of Things applications. Although Internet of Things sensors are widely distributed， some remote areas cannot be covered by mobile networks. Introducing unmanned aerial vehicle （UAV） into wireless networks can enhance network coverage performance and increase spectrum utilization. However， wireless communication signals are easily eavesdropped on， which raises concerns regarding the potential leakage of user information. This study proposes a joint optimization algorithm for energy harvesting and backscatter communication security capacity to address the security issues associated with UAV-assisted backscatter communication systems. Under the constraints of the energy and security performance for backscatter communication， an optimization problem was constructed to maximize the minimum backscatter security capacity. As this optimization problem is non-convex， this study employed the block coordinate descent （BCD） global iterative algorithm to solve it， decomposing the original optimization problem into three optimization sub-problems. Linear programming was utilized to solve the time allocation sub-problem. For the non-convex UAV trajectory optimization sub-problem， the sub-problem was first simplified by substituting the variable and then solved using successive convex approximation （SCA）. Solving the non-convex ground user scheduling sub-problem necessitated the application of the branch and bound method as the optimization variables to be solved were integer variables. Finally， the global optimal solutions for time allocation， UAV trajectory， and ground backscatter equipment scheduling were obtained via global iteration. The simulation results demonstrated that the algorithm proposed in this study has good convergence. Compared with the other two reference algorithms listed， the algorithm proposed in this study had better performance and could effectively improve the security capacity of the system.