A5083300391- Neural Networks and Applications
- Neural Networks and Reservoir Computing
- Advanced Memory and Neural Computing
- Robotic Path Planning Algorithms
- Reinforcement Learning in Robotics
- Time Series Analysis and Forecasting
- Stock Market Forecasting Methods
- Metamaterials and Metasurfaces Applications
- Robotics and Automated Systems
- Photovoltaic System Optimization Techniques
- Electromagnetic wave absorption materials
- EEG and Brain-Computer Interfaces
- Advanced Antenna and Metasurface Technologies
- Solar Radiation and Photovoltaics
- Energy Load and Power Forecasting
- Face recognition and analysis
- Calcium Carbonate Crystallization and Inhibition
- Gear and Bearing Dynamics Analysis
- Neural dynamics and brain function
- Visual Attention and Saliency Detection
- Advanced Multi-Objective Optimization Algorithms
- Vehicle Routing Optimization Methods
- Evolutionary Algorithms and Applications
- Machine Fault Diagnosis Techniques
- Robotics and Sensor-Based Localization
Zhengzhou University
2016-2025
Yantai University
2022-2024
Ningbo University
2023
Ningbo Science and Technology Bureau
2023
North China Electric Power University
2009
Dynamic multimodal optimization problems (DMMOPs) represent the that optimal solution changes over time. Due to wide application of DMMOPs in reality, some related algorithms have been proposed recent years. Most existing employ a single dynamic response mechanism and embed it multi-modal evolutionary algorithms. However, these often perform limited when environmental change involves multiple types, they fail consider utilizing historical information assist static optimizers. To solve...
The current trajectory planning methods for multi‐robot systems face challenges due to high computational burden and inadequate adaptability in complex constrained environments, obstructing efficiency improvements production logistics. This article presents an innovative solution by integrating model predictive contouring control (MPCC) long short‐term memory (LSTM) networks real‐time of multiple mobile robots. Based on the datasets generated MPCC, a customized LSTM network is constructed...
As an efficient recurrent neural network (RNN), reservoir computing (RC) has achieved various applications in time-series forecasting. Nevertheless, a poorly explained phenomenon remains as to why the RC and deep RCs succeed handling prediction despite completely randomized weights. This study tries generate grouped vector autoregressive (GVARC) forecasting model based on randomly distributed embedding (RDE) theory. In RDE-GVARC, structures are constructed by multiple GVARCs, which makes...
Intractable sparse feature extraction, over-weight model size, and limited training samples are currently bewildering in infrared dim small target detection, which not adequately addressed by current state-of-the-art (SOTA) methods. Here, to synchronously address these issues, a dense multi-level extraction fusion network (DMEF-net) is designed, mainly consisting of two modules: context Gaussian saliency module (TCGS) structure (MLDF). Inspired the physical thermal diffusion human visual...
Multivariate time series (MTS) play essential roles in daily life because most real-world datasets are multivariate and rich time-dependent information. Traditional forecasting methods for MTS time-consuming filled with complicated limitations. One efficient method being explored within the dynamical systems is extended short-term memory networks (LSTMs). However, existing models only partially use hidden spatial relationship as effectively LSTMs. Shallow LSTMs inadequate extracting features...
Deep learning bearing-fault diagnosis has shown strong vitality in recent years. In industrial practice, the running state of bearings is monitored by collecting data from multiple sensors, for instance, drive end, fan and base. Given complexity operating conditions limited number samples, obtaining complementary fault features using traditional fault-diagnosis method, which uses statistical characteristic time or frequency, difficult relies heavily on prior knowledge. addition, intelligent...
Abstract Face recognition is one of the core and challenging issues in computer vision field. Compared to vision, human visual system can identify a target from complex backgrounds quickly accurately. This paper proposes new network model deriving Where-What Networks (WWNs), which approximately simulate information processing pathways (i.e., dorsal pathway ventral pathway) cortex recognize different types faces with locations sizes background. To enhance performance, synapse maintenance...
Online model-free reinforcement learning (RL) approaches play a crucial role in coping with the real-world applications, such as behavioral decision making robotics. How to balance exploration and exploitation processes is central problem RL. A balanced ratio of exploration/exploitation has great influence on total time quality learned strategy. Therefore, various action selection policies have been presented obtain between procedures. However, these are rarely, automatically, dynamically...
During the environmental cognition, how to realize efficient incremental learning of a mobile robot is great challenge. The existing methods suffer from low efficiency. This article proposes novel methodology address it. A memory model inspired by human brains constructed transmission short-term (STM) long-term (LTM) in offline states, thus robot. Concretely, during online process, when sensory information input developmental network (DN), similarity between new and knowledge memorized DN...