采用D-K-D迭代算法设计导弹增益调度自动驾驶仪

Missile gain-scheduled autopilot design using D-K-D iteration

导弹增益调度自动驾驶仪除具有对时变参数的自适应调节能力外,还应具有对不可测量不确定性的抑制能力.为此,将导弹自动驾驶仪设计问题描述为一类含有不可测量不确定性的线性参数时变(LPV)系统的鲁棒增益调度问题,构造了模型匹配自动驾驶仪设计结构,并通过D–K–D迭代算法综合运用LPV控制方法和μ综合方法设计了导弹鲁棒增益调度自动驾驶仪.设计的自动驾驶仪不仅能够随导弹飞行马赫数和高度的变化自动进行参数调节,还能够有效抑制量测噪声、量测误差及建模误差等不可测量不确定性.仿真结果验证了设计方法的有效性和可行性.

Missile gain-scheduled autopilot should not only adapt to the variations of time-varying parameters, but also restrain unmeasured uncertainties. By describing missile autopilot design problem as a robust gain-scheduled problem of linear parameter-varying（LPV） system with unmeasured uncertainties, we establish a model matching autopilot design structure, and a missile robust gain-scheduled autopilot is then designed by combining the LPV control approach and the μ-synthesis approach through the D-K-D iteration. The designed autopilot not only adjusts parameters automatically with respect to the flight Mach number and altitude, but also restrains unmeasured uncertainties such as measurement noise, measurement errors and modeling errors. Simulation results validate the effectiveness and feasibility of the approach.