Search

Intelligent video analytics (IVA) plays an increasingly important role in industrial applications such as autonomous driving, defect detection and logistic scene analytics. Semantic segmentation is a basic technique of IVA, as well as lightweight and efficient segmentation is especially significant in industrial applications. But most excellent methods focus on accuracy, rather than lightweight and fast. Therefore, we propose a unique lightweight and efficient semantic segmentation network aimed at logistics truck scene analysis. Firstly, the lightweight high-level semantic feature extractor, which is a deep branch, is proposed to get a larger receptive field with lower parameters. Secondly, the detail block as a shallow branch is designed to enrich the spatial detail information. Lastly, the feature guided aggregation module based on the novel two-branch methods proposed to guide feature fusion for improving the performance. The parameters of our model are only 1.05M, which are much less than other models, such as PSPNet, SegNet and BiseNet. Compared with some typical real-time segmentation methods, our approach achieves more competitive accuracy (83.94%) and higher inference speed (157 FPS) in the logistics truck scene dataset.

Soft continuum manipulators provide a safe alternative to traditional rigid manipulators, because their bodies can absorb and distribute contact forces. Soft manipulators have near infinite potential degrees of freedom, but a limited number of control inputs. This underactuation means soft continuum manipulators often lack either the controllability or the dexterity to achieve desired tasks. In this work, we present an extension of McKibben actuators, which have well-known models, that increases the controllable degrees of freedom using active reconfiguration of the constraining fibers. These Active Fiber Reinforced Elastomeric Enclosures (AFREEs) preform some combination of length change and twisting, depending on the fiber configuration. Experimental results shows that by changing the fiber angles within a range of -30 to 30 degrees and actuating the resulting configuration between 10.3 kPa and 24.1 kPa, we can achieve twists between ± 60 degrees and displacements between -2 and 4 mm. By additionally controlling the fiber lengths and pressure, we can modify the AFREE kinematics further, creating dynamic behaviors and trajectories of actuation. The presented actuator creates the possibility to reconFigure actuator kinematics to meet desired soft robot motions.

With photovoltaic technology development and application of the expansion of photovoltaic, the accuracy of power generation forecast will directly affect the grid load frequency, grid stability, system reliability, power quality, investment and the system operating efficiency. The short-term forecasting research of photovoltaic power generation will help to stabilize power system operation, to optimize grid dispatching and to provide effective protection for the distribution load management. The rough sets theory was first proposed to the forecast of photovoltaic power generation and given the forecast mathematical models in this paper. The redundant attribute was removed from the decision table with the attribute reduction algorithm of Pawlak attribute importance. The continuous attributes of PV systems is discrete by the improved greedy algorithm. Results comparison of actual operation data and the simulation of week, month, season and year shows that rough sets forecast models is more correct and accuracy.

A novel control strategy of Fuzzy-Immune PID based on GA is proposed in this paper. The strategy learns from the immune feedback principle of biological immune system and takes advantage of the good approaching ability of fuzzy control to achieve nonlinear function of the immune controller, and then the GA algorithm is used to optimize parameters of PID controller and immune controller. Apply the strategy into the bending control system of rolling mills with characteristics of high nonlinearity, time-varying and time-delay. The simulation results indicate that the strategy has strong adaptability to the transformation of the system parameters and has advantages of low overshoot, short adjusting time, strong anti-disturbance ability and great robustness.

In view of the characteristics of nonlinearity, time-varying and time-delay in hydraulic AGC control system, a Smith fuzzy-PID complex control strategy is proposed. Smith predictor algorithm is used to compensate for the time delay and fuzzy control is used to enhance the robustness of the system, and then PID algorithms are used to improve the static precision of the system. The simulation results indicate that the method has characteristics of small overshoot, short adjusting time and strong anti-interference ability and robustness. It can further improve the quality of the strip thickness.