In this paper, a robust output-feedback adaptive control is proposed for linear time-invariant (LTI) single-input single-output (SISO) plants with unmeasurable input disturbance. Using dynamic surface control (DSC) technique, it is shown that the explosion of complexity problem in backstepping control can be eliminated. Furthermore, the proposed adaptive DSC scheme has the following merits: 1) by introducing an initialization technique, the L ∞ performance of system tracking error can be guaranteed even if the plant high-frequency gain is unknown and the input disturbance exists, and 2) the adaptive law is necessary only at the first design step, which significantly reduces the design procedure. It is proved that with the proposed scheme, all the closed-loop signals are semiglobally uniformly ultimately bounded. Simulation results are presented to demonstrate the effectiveness of the proposed scheme.
针对现有文本检测与定位方法只能处理单一方向文本行的缺点,提出了一种基于语义分割方法的用于自然图像中文本检测的新方法。首先通过对现有检测方法以及目前语义分割方法在文本行检测中的局限性分析。然后对加入矩形卷积核的全卷积网络模型进行训练,获得文本行区域的分类图。最后,通过全连接条件随机场(conditional random field,CRF)的高精度分割能力将网络前端输出的文本行区域中的文字给区分出来。该框架用于处理任意方向、语言和字体中的文本。所提出的方法在MSRA-TD500和ICDAR2015两个文本检测数据集上获得良好的分割结果且性能优越。
针对多模型自适应估计(multiple model adaptive estimation,MMAE)方法适应突变故障能力差、多重渐消因子强跟踪算法滤波发散、故障条件概率计算量大等问题,提出一种改进的多重渐消因子强跟踪多模型自适应估计(strong tracking multiple model adaptive estimation,STMMAE)快速故障诊断方法。通过多重渐消因子提高了故障突变时滤波器的跟踪性能;通过改进一步预测协方差阵更新方程,保证了滤波器稳定性,提高了估计精度;采用基于欧几里得范数的飞机舵面故障概率快速计算方法,降低了故障概率计算量。对比仿真表明,该算法跟踪性强、速度快、精度高,具有较好的鲁棒性和稳定性。
