Locomotion Control With Frequency and Motor Pattern Adaptations

Existing adaptive locomotion control mechanisms for legged robots are usually aimed at one specific type of adaptation and rarely combined with others. Adaptive mechanisms thus stay at a conceptual level without their coupling effect with other mechanisms being investigated. However, we hypothesize that the combination of adaptation mechanisms can be exploited for enhanced and more efficient locomotion control as in biological systems. Therefore, in this work, we present a central pattern generator (CPG) based locomotion controller integrating both a frequency and motor pattern adaptation mechanisms. We use the state-of-the-art Dual Integral Learner for frequency adaptation, which can automatically and quickly adapt the CPG frequency, enabling the entire motor pattern or output signal of the CPG to be followed at a proper high frequency with low tracking error. Consequently, the legged robot can move with high energy efficiency and perform the generated locomotion with high precision. The versatile state-of-the-art CPG-RBF network is used as a motor pattern adaptation mechanism. Using this network, the motor patterns or joint trajectories can be adapted to fit the robot's morphology and perform sensorimotor integration enabling online motor pattern adaptation based on sensory feedback. The results show that the two adaptation mechanisms can be combined for adaptive locomotion control of a hexapod robot in a complex environment. Using the CPG-RBF network for motor pattern adaptation, the hexapod learned basic straight forward walking, steering, and step climbing. In general, the frequency and motor pattern mechanisms complement each other well and their combination can be seen as an essential step toward further studies on adaptive locomotion control.

Equipment Maintenance Support Effectiveness Evaluation Based on Improved Generative Adversarial Network and Radial Basis Function Network

Due to the lack of maintenance support samples, maintenance support effectiveness evaluation based on the deep neural network often faces the problem of small sample overfitting and low generalization ability. In this paper, a neural network evaluation model based on an improved generative adversarial network (GAN) and radial basis function (RBF) network is proposed to amplify maintenance support samples. It adds category constraint based on category probability vector reordering function to GAN loss function, avoids the simplification of generated sample categories, and enhances the quality of generated samples. It also designs a parameter initialization method based on parameter components equidistant variation for RBF network, which enhances the response of correct feature information and reduces the risk of training overfitting. The comparison results show that the mean square error (MSE) of the improved GAN-RBF model is 5.921 × 10 − 4 , which is approximately 1/2 of the RBF model, 1/3 of the Elman model, and 1/5 of the BP model, while its complexity remains at a reasonable level. Compared with traditional neural network evaluation methods, the improved GAN-RBF model has higher evaluation accuracy, better solves the problem of poor generalization ability caused by insufficient training samples, and can be more effectively applied to maintenance support effectiveness evaluation. At the same time, it also provides a good reference for evaluation research in other fields.

Neural Network Optimization and Data Fusion Recognition Method for Intelligent Mechanical Fault Diagnosis

With the improvement of mechanical equipment complexity and automation level, the importance of mechanical equipment fault diagnosis is more and more prominent, and the choice of appropriate diagnosis method is crucial to the accuracy of the diagnosis results. Wavelet analysis and neural network technology, as the hot spot and frontier of research, are also important research contents in the development of intelligent diagnosis of mechanical fault. Data fusion can process multisource information to obtain more accurate and reliable methods. At the same time, because of its good nonlinearity, adaptability, and fault tolerance, neural network has become the preferred method of mechanical fault diagnosis. This paper first describes the research content and significance of fault diagnosis technology and introduces the main methods and steps of fault diagnosis, and through the introduction of mechanical fault vibration signals, vibration signals were analyzed in time domain and frequency domain. Secondly, the definition and classification of data I fusion and RBF neural network are introduced in detail and compared with BP neural network. Because the prediction accuracy of the RBF network is higher than that of the BP neural network and the training time of the RBF network is obviously shorter than that of the BP network, the RBF network has significant advantages over diagnostic errors. In this paper, six valve signals were collected under normal conditions and errors, and by analyzing and comparing different theoretical foundations, the 4-second network crisis time was effectively reduced, which provided the basis for teaching monitoring.

Adaptive Parameter Modulation of Deep Brain Stimulation Based on Improved Supervisory Algorithm

Clinically deployed deep brain stimulation (DBS) for the treatment of Parkinson’s disease operates in an open loop with fixed stimulation parameters, and this may result in high energy consumption and suboptimal therapy. The objective of this manuscript is to establish, through simulation in a computational model, a closed-loop control system that can automatically adjust the stimulation parameters to recover normal activity in model neurons. Exaggerated beta band activity is recognized as a hallmark of Parkinson’s disease and beta band activity in model neurons of the globus pallidus internus (GPi) was used as the feedback signal to control DBS of the GPi. Traditional proportional controller and proportional-integral controller were not effective in eliminating the error between the target level of beta power and the beta power under Parkinsonian conditions. To overcome the difficulties in tuning the controller parameters and improve tracking performance in the case of changes in the plant, a supervisory control algorithm was implemented by introducing a Radial Basis Function (RBF) network to build the inverse model of the plant. Simulation results show the successful tracking of target beta power in the presence of changes in Parkinsonian state as well as during dynamic changes in the target level of beta power. Our computational study suggests the feasibility of the RBF network-driven supervisory control algorithm for real-time modulation of DBS parameters for the treatment of Parkinson’s disease.

Integrating Sponge City Concept and Neural Network into Land Suitability Assessment: Evidence from a Satellite Town of Shenzhen Metropolitan Area

Land suitability assessment is fundamental in space control planning and land development because of its effects on land use and urban layout. Rainstorms and waterlogging have become one of the most common natural disasters in the coastal areas of China. As a result, the concept of an ecological sponge city was incorporated into the construction of cities in the future. Taking Shenzhen–Shantou special cooperation zone (SSCZ), we constructed a storm flooding model based on the SCS flow generation model and GIS to explore the spatial distribution characteristics of the flooding risk in a rainstorm of 100-year lasting 1 h. Combined with population and economic indicators, a radial basis function (RBF) network was utilized to evaluate the environmental risk, the vulnerability of disaster-bearing bodies, and the rain–flood resilience of sponge cities. The self-organizing feature mapping (SOFM) model was used for cluster analysis. Spatial differences were found in the construction suitability of the study area. A suitable construction area (73.59% of the entire area) was located downtown. The construction of the artificial spongy body in the highest vulnerable area (3.25%) needs to be strengthened. The control construction area (3.3%) is located along the banks of the river, with relatively high risk and low resilience of flood control engineering. Ecological construction (19.85%) serves as the sponge body of ecological buffer. The factors of waterlogging, ecology, population, and economy could be integrated comprehensively by applying neural network methods for urban planning and construction.

Improved RBF Network Intrusion Detection Model Based on Edge Computing with Multi-algorithm Fusion

Edge computing is difficult to deploy a complete and reliable security strategy due to its distributed computing architecture and inherent heterogeneity of equipment and limited resources. When malicious attacks occur, the loss will be immeasurable. RBF neural network has strong nonlinear representation ability and fast learning convergence speed, which is suitable for intrusion detection of edge detection industrial control network. In this paper, an improved RBF network intrusion detection model based on multi-algorithm fusion is proposed. kernel principal component analysis (KPCA) is used to extract data dimension and simplify data representation. Then subtractive clustering algorithm(SCM) and grey wolf algorithm(GWO) are used to jointly optimize RBF neural network parameters to avoid falling into local optimum, reduce the calculation of model training and improve the detection accuracy. The algorithm can better adapt to the edge computing platform with weak computing ability and bearing capacity, and realize real-time data analysis.The experimental results of BATADAL data set and Gas data set show that the accuracy of the algorithm is over 99% and the training time of larger samples is shortened by 50 times for BATADAL data set. The results show that the improved RBF network is effective in improving the convergence speed and accuracy in intrusion detection.

Integrated neural network model with pre-RBF kernels

To improve the network performance of radial basis function (RBF) and back-propagation (BP) networks on complex nonlinear problems, an integrated neural network model with pre-RBF kernels is proposed. The proposed method is based on the framework of a single optimized BP network and an RBF network. By integrating and connecting the RBF kernel mapping layer and BP neural network, the local features of a sample set can be effectively extracted to improve separability; subsequently, the connected BP network can be used to perform learning and classification in the kernel space. Experiments on an artificial dataset and three benchmark datasets show that the proposed model combines the advantages of RBF and BP networks, as well as improves the performances of the two networks. Finally, the effectiveness of the proposed method is verified.