针对预测-校正内点法(predictor-corrector primal-dualinterior point method,PCPDIPM)加权最小绝对值状态估计(weighted least absolute squares,WLAV)可能发生校正方向指向错误方向的不足,提出一种基于多预测-校正内点法(multiple PC...针对预测-校正内点法(predictor-corrector primal-dualinterior point method,PCPDIPM)加权最小绝对值状态估计(weighted least absolute squares,WLAV)可能发生校正方向指向错误方向的不足,提出一种基于多预测-校正内点法(multiple PCPDIPM,MPCPDIPM)的WLAV抗差状态估计算法。该算法在PCPDIPM的基础上,通过多次校正,对中心参数动态估计,并采用2阶段线性搜索法确定校正方向在总的牛顿方向中的最优比重,从而保证迭代点向中心轨迹靠拢。最后,通过IEEE算例仿真和我国某省网的测试结果验证了所提方法的有效性。与含不良数据辨识功能的加权最小二乘状态估计相比较,所提方法的收敛速度及抗差能力具有明显的优势。展开更多
A skip entry guidance algorithm blending numerical predictor-corrector and nominal trajectory tracking is presented for lunar return vehicles.The guidance is decoupled into longitudinal and lateral channels.A piecewis...A skip entry guidance algorithm blending numerical predictor-corrector and nominal trajectory tracking is presented for lunar return vehicles.The guidance is decoupled into longitudinal and lateral channels.A piecewise bank-vs-energy magnitude profile and a sign profile are adopted in the skip phase.A magnitude parameter is used to adjust the predicted downrange,and a pseudo-crossrange at the beginning of the final phase is selected as the lateral control variable.Prediction biases of both channels are nullified by a false position iteration algorithm.An on-line estimation and modeling method is introduced to compensate for aerodynamic and atmospheric uncertainties.A nominal trajectory for the final phase is generated based on actual reenter conditions,and the obtained nominal trajectory is tracked by a linear feedback law.A lateral corridor is used to manage the lateral state.The proposed guidance algorithm is assessed using three-degree-of-freedom Monte Carlo analyses,and the results show a satisfactory and robust performance under highly stressful dispersions.展开更多
文摘针对预测-校正内点法(predictor-corrector primal-dualinterior point method,PCPDIPM)加权最小绝对值状态估计(weighted least absolute squares,WLAV)可能发生校正方向指向错误方向的不足,提出一种基于多预测-校正内点法(multiple PCPDIPM,MPCPDIPM)的WLAV抗差状态估计算法。该算法在PCPDIPM的基础上,通过多次校正,对中心参数动态估计,并采用2阶段线性搜索法确定校正方向在总的牛顿方向中的最优比重,从而保证迭代点向中心轨迹靠拢。最后,通过IEEE算例仿真和我国某省网的测试结果验证了所提方法的有效性。与含不良数据辨识功能的加权最小二乘状态估计相比较,所提方法的收敛速度及抗差能力具有明显的优势。
基金supported by the National Natural Science Foundation of China(61203194)the Innovation Fund of National University of Defense Technology(B100101)
文摘A skip entry guidance algorithm blending numerical predictor-corrector and nominal trajectory tracking is presented for lunar return vehicles.The guidance is decoupled into longitudinal and lateral channels.A piecewise bank-vs-energy magnitude profile and a sign profile are adopted in the skip phase.A magnitude parameter is used to adjust the predicted downrange,and a pseudo-crossrange at the beginning of the final phase is selected as the lateral control variable.Prediction biases of both channels are nullified by a false position iteration algorithm.An on-line estimation and modeling method is introduced to compensate for aerodynamic and atmospheric uncertainties.A nominal trajectory for the final phase is generated based on actual reenter conditions,and the obtained nominal trajectory is tracked by a linear feedback law.A lateral corridor is used to manage the lateral state.The proposed guidance algorithm is assessed using three-degree-of-freedom Monte Carlo analyses,and the results show a satisfactory and robust performance under highly stressful dispersions.
