【摘要】 为解决车载反无人机火炮在复杂扰动下控制性能不足的问题，设计了一种融合BP神经网络与离散时间重复控制的位置环自抗扰策略(BPNN-DRC-ADRC)。该方法通过BP神经网络实时调整控制器参数以提升动态响应，采用混沌粒子群算法(CPSO)对神经网络的权值进行寻优，优化控制器性能，并利用离散时间重复控制器(DRC)抑制车体的周期性扰动以提高稳态精度，使得参数协同优化。仿真与实验表明，该策略能有效提升系统跟踪时的控制性能。更多还原

【Abstract】 To address the insufficient control performance of vehicle-mounted anti-drone artillery under complex disturbances, a position-loop active disturbance rejection control strategy integrating a BP neural network and discrete-time repetitive control(BPNN-DRC-ADRC) is proposed. This method utilizes a BP neural network to adjust controller parameters in real-time to enhance dynamic response, while a chaotic particle swarm optimization(CPSO) algorithm is employed to optimize the weights of the neural network for improved controller performance. Furthermore, a discrete-time repetitive controller(DRC) is incorporated to suppress periodic disturbances from the vehicle body to increase steady-state precision, achieving collaborative parameter optimization. Simulation and experimental results demonstrate that this strategy effectively improves the control performance of the system during tracking tasks.更多还原