Based on the method of unidirectional causality measure, this paper analyzes the long-term and short-term dynamic effects and causality between China’s population aging and technological innovation. According to the empirical results, first, the aging of the population will eventually have a continuous long-term impact, although it has little effect on the technology innovation in the short term. Second, when compared with the old-age dependency ratio, the child-raising ratio has a remarkable unidirectional causal effect on the technological innovation in the short term. Third, when compared with the old-age dependency ratio, the total dependency ratio has a stronger impact on the scientific and technological innovation ability, which is a long-term effect. The finding indicates that the elderly population and the children’s population have a continuous impact on China’s scientific and technological innovation, that is, the increase in social support burden affects the technological innovation for a long time.
Traditional convolutional neural networks (CNNs) use a pooling layer to reduce the dimensionality of texts, but lose semantic information. To solve this problem, this paper proposes a convolutional neural network model based on singular value decomposition algorithm (SVD-CNN). First, an improved density-based center point clustering active learning sampling algorithm (DBC-AL) is used to obtain a high-quality training set at a low labelling cost. Second, the method uses the singular value decomposition algorithm for feature extraction and dimensionality reduction instead of a pooling layer, fuses the dimensionality reduction matrix, and completes the barrage text classification task. Finally, the partial sampling gradient descent algorithm (PSGD) is applied to optimize the model parameters, which accelerates the convergence speed of the model while ensuring stability of the model training. To verify the effectiveness of the improved algorithm, several barrage datasets were used to compare the proposed model and common text classification models. The experimental results show that the improved algorithm preserves the semantic features of the text more successfully, ensures the stability of the training process, and improves the convergence speed of the model. Further, the model’s classification performance on different barrage texts is superior to traditional algorithms.
This paper aims to analyze the velocity pattern of a power-assisted mobile robot when the operator performs operation without any discomfort. Power-assist systems for mobile robots such as wheelchairs and conveyance carriers are extremely effective in alleviating the physical burden on operators when they carry heavy objects. Although the velocity control based power-assist system has an advantage that it can be easily realized, the problem lies in that the system becomes unstable when the operator has high stiffness. Variable impedance control based on impedance estimation of the operator is effective at solving this problem. To realize operator impedance estimation, it is necessary to know the intended robot’s motion of a person. In this study, as a preliminary step to estimate the operator’s impedance, the velocity pattern when the operator performs natural operation of the robot through the power-assist system is analyzed. The results confirm that the natural velocity pattern can be approximated by a velocity pattern connecting two minimum jerk trajectories.
When a nuclear power disaster occurs at a nuclear power plant, it is hazardous for humans to enter the plant. If robots could remove radioactive substances adhering to a plane such as a plant wall, humans would be able to enter the plant to investigate the situation and to work. In this study, to efficiently remove radioactive substances from a wall with a manipulator, we examined joint trajectory planning based on the minimum Euclidean distance of joint angles of a seven-degrees-of-freedom (7-DOF) serial link manipulator for a sequential reaching task on a plane. We demonstrate the planning for the sequential reaching task, which is an iterative point-to-point reaching movement between positions on a plane. The joint angles for each target position were obtained based on the inverse kinematics for an arm angle, and the optimal arm angles within the constraints of the joint angles were computed by the sequential quadratic programming method. The optimal trajectories for the arm angles were compared with the trajectories of the joint angles that were the eight inverse kinematic solutions for a fixed arm angle. The result showed that through optimal planning, an efficient trajectory within the movable ranges of the joint angles could be obtained for the sequential reaching task.
A novel growth evaluation system for tobacco planting (GESTP) based on a B/S architecture is introduced in this paper. It mainly consists of three parts: a mobile application (mobile app), a browser terminal and a server terminal. The GESTP system is used to evaluate the growth of tobacco and give farmers planting guidance instead them having to rely on personal judgment. Once the photos of the tobacco leaf and plant are uploaded to the web server via the mobile app or the browser terminal, the application program of the server terminal is called to process the tobacco images with image processing algorithms. The results including the grade of the tobacco growth and planting guidance will be provided to the client within a 2-second timeframe, which greatly help farmers understand the growth of tobacco and take planting measures. The running result indicates that the GESTP system provides an effective and straightforward way to evaluate the growth of tobacco and provides cultivation guidance to tobacco farmers.
Optical devices often achieve their maximum effectiveness by using dielectric mirrors; however, their design techniques depend on expert knowledge in specifying the mirror properties. This expertise can also be achieved by machine learning, although it is not clear what kind of neural network would be effective for learning about dielectric mirrors. In this paper, we clarify that the recurrent neural network (RNN) is an effective approach to machine-learning for dielectric mirror properties. The relation between the thickness distribution of the mirror’s multiple film layers and the average reflectivity in the target wavelength region is used as the indicator in this study. Reflection from the dielectric multilayer film results from the sequence of interfering reflections from the boundaries between film layers. Therefore, the RNN, which is usually used for sequential data, is effective to learn the relationship between average reflectivity and the thickness of individual film layers in a dielectric mirror. We found that a RNN can predict its average reflectivity with a mean squared error (MSE) less than 10-4 from representative thickness distribution data (10 layers with alternating refractive indexes 2.3 and 1.4). Furthermore, we clarified that training data sets generated randomly lead to over-learning. It is necessary to generate training data sets from larger data sets so that the histogram of reflectivity becomes a flat distribution. In the future, we plan to apply this knowledge to design dielectric mirrors using neural network approaches such as generative adversarial networks, which do not require the know-how of experts.
The difficulties in implementing the model predictive control (MPC) in interior permanent-magnet synchronous motors (IPMSMs) consist of the nonlinear behavior of IPMSMs and the computational effort required by MPC. This paper presents an IPMSM controller design method for electric vehicles based on explicit MPC (EMPC), which uses a different linearization method. The proposed controller combines the speed and current controllers and replaces the traditional cascade structure. First, the nonlinear terms in the system model are added into the control input as voltage compensation to obtain a simple linear model. Next, the proposed controller based on MPC is designed, which considers the effects of load torque and uses an increment model. Furthermore, the controller applies both current and voltage constraints. The EMPC method based on a binary search is used to accelerate the solution of the optimization problem. Finally, the simulation results show the validity and superiority of the proposed method.
A method is proposed for reducing noise in learning data based on fuzzy inference methods called α-GEMII (α-level-set and generalized-mean-based inference with the proof of two-sided symmetry of consequences) and α-GEMINAS (α-level-set and generalized-mean-based inference with fuzzy rule interpolation at an infinite number of activating points). It is particularly effective for reducing noise in randomly sampled data given by singleton input–output pairs for fuzzy rule optimization. In the proposed method, α-GEMII and α-GEMINAS are performed with singleton input–output rules and facts defined by fuzzy sets (non-singletons). The rules are initially set by directly using the input–output pairs of the learning data. They are arranged with the facts and consequences deduced by α-GEMII and α-GEMINAS. This process reduces noise to some extent and transforms the randomly sampled data into regularly sampled data for iteratively reducing noise at a later stage. The width of the regular sampling interval can be determined with tolerance so as to satisfy application-specific requirements. Then, the singleton input–output rules are updated with consequences obtained in iteratively performing α-GEMINAS for noise reduction. The noise reduction in each iteration is a deterministic process, and thus the proposed method is expected to improve the noise robustness in fuzzy rule optimization, relying less on trial-and-error-based progress. Simulation results demonstrate that noise is properly reduced in each iteration and the deviation in the learning data is suppressed considerably.
In this work, we propose a multi-channel semantic fusion convolutional neural network (SFCNN) to solve the problem of emotional ambiguity caused by the change of contextual order in sentiment classification task. Firstly, the emotional tendency weights are evaluated on the text word vector through the improved emotional tendency attention mechanism. Secondly, the multi-channel semantic fusion layer is leveraged to combine deep semantic fusion of sentences with contextual order to generate deep semantic vectors, which are learned by CNN to extract high-level semantic features. Finally, the improved adaptive learning rate gradient descent algorithm is employed to optimize the model parameters, and completes the sentiment classification task. Three datasets are used to evaluate the effectiveness of the proposed algorithm. The experimental results show that the SFCNN model has the high steady-state precision and generalization performance.
Cognitive radio employs an opportunistic spectrum access approach to ensure efficient utilization of the available spectrum by secondary users (SUs). To allow SUs to access the spectrum opportunistically, the spectrum sensing process must be fast and accurate to avoid possible interference with the primary users. Previously, two-stage spectrum sensing methods were proposed that consider the sensing time and sensing accuracy parameters independently at the cost of a non-optimal spectrum sensing performance. To resolve this non-optimality issue, we consider both parameters in the design of our spectrum sensing scheme. In our scheme, we first derive optimal thresholds using an optimization equation with an objective function of maximizing the probability of detection, subject to the minimal probability of error. We then minimize the average spectrum sensing time using signal-to-noise ratio estimation. Our simulation results show that the proposed improved two-stage spectrum sensing (ITSS) scheme provides a 4%, 7%, and 6% better probability of detection accuracy rate than two-stage combinations of energy detection (ED) and maximum eigenvalue detection, energy detection and cyclostationary feature detection (CFD), and ED and combination of maximum-minimum eigenvalue (CMME) detection, respectively. The ITSS is superior also to single-stage ED by 19% and shows an improved average spectrum sensing time.
Current and future wireless network architectures consist of several access technologies to support numerous traffic types and enable mobile devices to be connected anytime, anywhere. However, providing a rapid seamless connectivity and service continuity between such various access technologies remains a challenge. This is mainly because the previously proposed handover algorithms have failed to predict the future values of the measured received signal strength needed for rapid handover process. In addition, existing handover algorithms are not adaptable to the changes of the network conditions and user preferences. This leads to erroneous network selection, packet loss, and ping-pong effect due to high-ranking abnormality. In this study, an intersystem handover (IH) algorithm has been designed by integrating grey prediction theory, multiple-attribute decision making, fuzzy analytic hierarchy process, and multi-objective optimization ratio analysis. Network Simulator 2 has been applied to evaluate the performance of the proposed IH algorithm when compared to the fuzzy logic-based vertical handover (FLBVH) algorithm and the adaptive neuro-fuzzy inference system (ANFIS) algorithm. On average, the proposed IH algorithm has shown 1.1 s handover delay, 5% packet loss, 1.6% probability of ping-pong effect, and 97.8% better throughput performance than the ANFIS algorithm and FLBVH algorithm, respectively, for a 100-s time interval.
To address problems relating to microscopic micro-vessel images of living bodies, including poor vessel continuity, blurry boundaries between vessel edges and tissue and uneven field illuminance, and this paper put forward a fuzzy-clustering level-set segmentation algorithm. By this method, pre-treated micro-vessel images were segmented by the fuzzy c-means (FCM) clustering algorithm to obtain original contours of interesting areas in images. By the evolution equations of the improved level set function, accurate segmentation of microscopic micro-vessel images was realized. This method can effectively solve the problem of manual initialization of contours, avoid the sensitivity to initialization and improve the accuracy of level-set segmentation. The experiment results indicate that compared with traditional micro-vessel image segmentation algorithms, this algorithm is of high efficiency, good noise immunity and accurate image segmentation.
Training tactile sensing for shape recognition is considered to be an effective rehabilitation technique. Previous studies in tactile sensing showed a tendency of recognition ambiguity, thus necessitating tactile sensing rehabilitation. Eleven subjects observed invisible objects using their fingers and were asked to identify the shape of the objects. The relationship between the degree of recognition and shape complexity was investigated. The results showed high self-confidence in recognizing high complexity shapes. The recognition process was confirmed in a second experiment measuring brain activation using near-infrared spectroscopy. Measurement of eight subjects showed the activation of verbal and visual processing regions, indicating that the act of handling the shape was translated to verbal expression and visual imaging. These results potentially quantify tactile sensing and contribute to the realization of personalized rehabilitation.
In our daily life, it is inevitable to confront the condition which we feel confident or unconfident. Under these conditions, we might have different expressions and responses. Not to mention under the situation when a human communicates with a robot. It is necessary for robots to behave in various styles to show adaptive confidence degree, for example, in previous work, when the robot made mistakes during the interaction, different certainty expression styles have shown influence on humans’ truthfulness and acceptance. On the other hand, when human feel uncertain on the robot’s utterance, the approach of how the robot recognizes human’s uncertainty is crucial. However, relative researches are still scarce and ignore individual characteristics. In current study, we designed an experiment to obtain human verbal and non-verbal features under certain and uncertain condition. From the certain/uncertain answer experiment, we extracted the head movement and voice factors as features to investigate if we can classify these features correctly. From the result, we have found that different people had distinct features to show different certainty degree but some participants might have a similar pattern considering their relatively close psychological feature value. We aim to explore different individuals’ certainty expression patterns because it can not only facilitate humans’ confidence status detection but also is expected to be utilized on robot side to give the proper response adaptively and thus spice up the Human-Robot Interaction.