Human pose estimation is an important research direction in the field of computer vision, which aims to accurately identify the position and posture of keypoints of the human body through images or videos. However, multi-person pose estimation yields false detection or missed detection in dense crowds, and it is still difficult to detect small targets. In this paper, we propose a Mamba-based human pose estimation. First, we design a GMamba structure to be used as a backbone network to extract human keypoints. A gating mechanism is introduced into the linear layer of Mamba, which allows the model to dynamically adjust the weights according to the different input images to locate the human keypoints more precisely. Secondly, GMamba as the backbone network can effectively solve the long-sequence problem. The direct use of convolutional downsampling reduces selectivity for different stages of information flow. We used slice downsampling (SD) to reduce the resolution of the feature map to half the original size, and then fused local features from four different locations. The fusion of multi-channel information helped the model obtain rich pose information. Finally, we introduced an adaptive threshold focus loss (ATFL) to dynamically adjust the weights of different keypoints. We assigned higher weights to error-prone keypoints to strengthen the model’s attention to these points. Thus, we effectively improved the accuracy of keypoint identification in cases of occlusion, complex background, etc., and significantly improved the overall performance of attitude estimation and anti-interference ability. Experimental results showed that the AP and AP50 of the proposed algorithm on the COCO 2017 validation set were 72.2 and 92.6. Compared with the typical algorithm, it was improved by 1.1% on AP50. The proposed method can effectively detect the keypoints of the human body, and provides stronger robustness and accuracy for the estimation of human posture in complex scenes.

In reality, wind power data are often accompanied by data losses, which can affect the accurate prediction of wind power and subsequently impact the real-time scheduling of the power system. Existing methods for recovering missing data primarily consider the environmental conditions of individual wind farms, thereby overlooking the spatiotemporal correlations between neighboring wind farms, which significantly compromise their recovery effectiveness. In this paper, a joint missing data recovery model based on power data from adjacent wind farms is proposed. At first, a spatial–temporal module (STM) is designed using a combination of graph convolution network and recurrent neural networks to learn spatiotemporal dependencies and similarities. Subsequently, to provide a solid computational foundation for the STM, a Euclidean-directed graph based on Granger causality is constructed to reflect the hidden spatiotemporal information in the data. Finally, comprehensive tests on data recovery for both missing completely at random and short-term continuous missing are conducted on a real-world dataset. The results demonstrate that the proposed model exhibits a significant advantage in missing data recovery compared to baseline models.

With the rapid development of the Internet of Things(IOT) and increasing research on tracking control of affine IOT systems with uncertain, a nonlinear control method of command filtering backstepping with self-organizing approximation is proposed. A command filter with a saturation structure is used to eliminate the analytic computation of command derivatives, and tracking errors are redefined to compensate the discrepancy between the command filtered signals and analytic computation of command derivatives. Backstepping command filtered is used to decouple the higher IOT system and simplify the structure of IOT system. Self-organizing approximation is used to on-line approximate the uncertain and eliminate the effect of the uncertain in IOT system. The approximation for the uncertain is finished by a linear product between a set of basis functions and weighted functions on the local region. The weighted sum of compensated tracking errors is used as the prespecified tracking accuracy performance. A new local approximator is added to the control system according to the criterion based on Lyapunov theory to enhance the capacity of the approximator when the prespecified tracking accuracy performance can’t be achieved. Finally, an application for a simple IOT system is analyzed by simulations. The simulation results show the control method for the IOT system is effective.