Yagang Wang’s research while affiliated with University of Shanghai for Science and Technology and other places

Passive brain-computer interfaces with electroencephalographs (EEGs) could measure human mental workload levels. However, the use of EEG headsets with varying electrode configurations presents challenges in creating an interpretable mental workload recognizer that can be applied across different individuals and databases. To address this issue, we propose a novel shared spatial map network (SSMN), which abstracts EEG data representations from different individuals and electrode layouts. The SSMN utilizes a shared map encoder to generate EEG feature maps with consistent spatial locations on the scalp. Then, an instance augmentation encoder is employed to increase the sample size and filter out individual-specific components. Finally, a workload recognition committee is established by combining shallow and deep architectures of convolutional neural networks, allowing for adaptation to different mental workload measurement scenarios. We evaluate SSMN’s performance by combining two feature types and four validation approaches on three databases. The classification accuracy of the mental workload level measured by the EEG samples is comparable with the state-of-the-art models. The shared feature maps also clearly interpret the contribution of the cortical regions to the mental workload variations.

When there are obstacles around the target point, the mobile robot cannot reach the target using the traditional artificial potential field (APF). Besides, the traditional APF is prone to local oscillation in complex terrain such as three‐point collinear or semiclosed obstacles. Aiming at solving the defects of traditional APF, a novel improved APF algorithm named back virtual obstacle setting strategy‐APF has been proposed in this paper. There are two main advantages of the proposed method. First, by redefining the gravitational function as a logarithmic function, the proposed method can make the mobile robot reach the target point when there are obstacles around the target. Second, the proposed method can avoid falling into local oscillation for both three‐point collinear and semiclosed obstacles. Compare with APF and other improved APF, the feasibility of the algorithm is proved through software simulation and practical application.

Recently, the application of convolution neural network (CNN) in single image super-resolution (SISR) is gradually developing. Although many CNN-based methods have acquired splendid performance, oversized model complexity hinders their application in real life. In response to this problem, lightweight and efficient are becoming development tendency of SR models. The residual feature distillation network (RFDN) is one of the state-of-the-art lightweight SR networks. However, the shallow residual block (SRB) in RFDN still uses ordinary convolution to extract feature, where still has great improvement room for the reduction of network parameters. In this paper, we propose the Group-convolutional Feature Enhanced Distillation Network (GFEDNet), which is constructed by the stacking of feature distillation and aggregation block (FDAB). Benefitting from residual learning of residual feature aggregation (RFA) framework and feature distillation strategy of RFDN, the FDAB can obtain more diverse and detailed feature representations, thereby improves the SR capability. Furthermore, we propose the multi-scale group convolution block (MGCB) to replace the SRB. Thanks to group convolution and multi-branch parallel structure, the MGCB reduces the parameters substantially while maintaining SR performance. Extensive experiments show the powerful function of our proposed GFEDNet against other state-of-the-art methods.

When there are obstacles around the target point, the mobile robot cannot reach the target using traditional Artificial Potential Field (APF). Besides, the traditional APF is prone to local oscillation in complex terrain such as three-point collinear or semi-closed obstacles. Aiming at solving the defects of traditional APF, a novel improved APF algorithm named back virtual obstacle setting strategy-APF (BVO-APF) has been proposed in this paper. There are two main advantages of the proposed method. Firstly, by redefining the gravitational function as logarithmic function, the proposed method can make the mobile robot reach the target point when there are obstacles around the target. Secondly, the proposed method can avoid falling into local oscillation for both three-point collinear and semi-closed obstacles. Compare with APF and other improved APF, the feasibility of the algorithm is proved through software simulation and practical application.

Aiming at the problem of object model identification of modern industrial process control systems, a new closed-loop moment parameter identification online method based on the data of normal operation of the running system is proposed. In this method, only one step response data of the system is required, and appropriate convergence factors are introduced into the Laplace formula, the trapezoidal integral method is used to calculate the values of two derivatives of the transfer function, then the four unknown parameters of the second-order model can be solved by fitting the data with the least square method, and the target model can be identified. Finally, the simulation results of building different objects through Matlab show that the identification method has general applicability and good robustness with high recognition, and it is not sensitive to noise signals.

Harris Hawks Optimizer (HHO) is a novel meta-heuristic algorithm that simulates the cooperative hunting behavior of the Harris hawks. Behind the good performance of HHO, there are still problems of poor solution accuracy and unavoidable local optimization. A new variant of HHO called NMHHO is proposed by introducing improved nonlinear escaping energy (E) and jump strength (J) in the original optimizer, and incorporating the mutation strategy of Difffferential Evolution algorithm (DE) for solving the global optimization. The improvement of nonlinear parameters improves the diversity of populations and ameliorates the imbalance between global exploration and local exploitation. The mutation strategy makes full use of population distribution information, and is very effffective for algorithm to jumping out of local optima. To verify the optimized performance of NMHHO, scalability analysis with difffferent dimensions and comparison with other swarm intelligence algorithms are performed on 23 classical benchmark functions derived from the original HHO. In addition, the actual optimization performance of the algorithm is evaluated on three practical engineering problems. Experimental results indicate that the improved NMHHO algorithm shows strong competitiveness, excellent optimization ability and robustness.

Aiming at the three-dimensional path planning of unmanned aerial vehicle (UAV) in the complex environment of material delivery in earthquake-stricken areas, this paper proposes an improved adaptive grey wolf optimization algorithm (AGWO) based on the grey wolf optimization algorithm (GWO). There are two main contributions of the proposed method. Firstly, we propose an adaptive convergence factor adjustment strategy and an adaptive weight factor to update the individual’s position. The effectiveness of the improved algorithm is verified by the convergence analysis and the test function simulation experiment. Secondly, the improved algorithm is applied to UAV path planning, the environmental map model is established by integrating digital elevation map and equivalent mountain threat model, and the performance evaluation function is established by fitting the calculated track length. The simulation results show that the improved AGWO is superior to the traditional intelligent algorithm in convergence precision, speed and stability performance, and it is effective for 3D trajectory optimization in complex environment.

In this paper, a novel tuning rule of differential forward proportional–integral–derivative (PID) controller for integrating systems with time delay is proposed by using the direct synthesis method. This method is based on comparing the characteristic equation of the system which consists of the differential forward PID controller adding a lead-lag filter and the integrating time-delay process with the desired characteristic equation. The desired characteristic equation is obtained by placing a plurality of poles at the same desired position. Parameters of the differential forward PID are tuned to achieve the desired robustness. Tuning rules are given in terms of plant parameters for various kinds of integrating systems. Compared with the conventional PID controller, the proposed method is more effective for frequently changed set-points. Simulations for various integrating processes as well as the non-linear jacketed CSTR system illustrate the applicability and effectiveness of the proposed method.