Applied Mechanics and Materials Vols. 303-306

Abstract: RBF neural network with the parameters randomly selected may have unstabilized error. In this paper on this problem we improve the RBFNN by introducing particle swarm optimization algorithm, and use the algorithm of searching optimization process to adjust the three parameters of RBFNN. Through simulating four groups of test data by improved PSO-RBFNN network, we get optimal parameters. The simulation shows that the error of the improved network is smaller.

Abstract: This paper absorbed research fruits of comprehensive information, emotional motive - information and three-dimensional PAD emotion theory, defined the main body of the priori knowledge and affection, and posteriori knowledge and affection, obtained knowledge and affections and expectation knowledge and affection, analyzed cognitive mechanism of knowledge, formation mechanism of affection, proposed knowledge-affection interaction mechanism model, deduced the principle of knowledge-affection interaction mechanism (that is, the first and second law), unveiled the human’s rich and complex knowledge-affection interaction and the emotion changing laws. Comprehensive cognitive emotional theory is a new doctrine of cognition - emotion interaction mechanism.

Abstract: This paper solves the overfitting phenomenon in neural network’s training process for improving the generalization ability of neural network ensemble system. It analyzes factors affecting the generalization ability theoretically and proposes methods of improving the generalization ability by research of existing methods and making individual network generation and improving training samples as the breakthrough.

Abstract: The multiple cases database construction is one of the important links to design the personalized recommendation system. Personalized recommendation system case can be organized with multiple cases database based on expert experience and thinking patterns, combined with the traditional case method of organization. This paper studies the multiple cases database construction method of the personalized recommendation system based-CBR.

Abstract: According to the subjectivity, randomness and fatigue of the satisfaction selection of online teaching evaluation indicators for the students, the level of teachers teaching can not truly measure by the evaluation results. This paper introduces the fuzzy comprehensive evaluation method to deal with the evaluation data, and Practice has proved that this method produced results more equitable to reflect on teaching level. Which help teachers to promptly improve and improve teaching, the variable quality of teaching time control for process control, change the static management for dynamic management.

Abstract: The Electric Vehicle Charging and Exchanging (EVCE) station is the important infrastructure for the Electric Vehicle (EV) industry. Meanwhile, the sensing technologies can effectively obtain the key parameters from kinds of smart devices, which make it very suitable for intelligent management in EVCE station system. In this paper, according to the environment and supporting technologies for EVCE station, we introduce the sensing technologies in detail, including technologies about EV battery sensing, EV sensing, charging/exchanging station sensing and charging pile sensing, etc. Practice shows that the sensing technology can effectively improve EVCE station intelligence.

Abstract: Building environments are significant sources of global energy consumption. To create energy efficient buildings, the first step is to sense and monitor all the energy-consuming appliances in the buildings and record all the energy consumption information. After that, appropriate energy saving policies can be decided and the instructions can be sent to the control devices to apply the energy saving adjustments. To do that, in-building two-way communication networks are needed to connect all the sensors to collect information as well as to send control instructions. However, most of the current devices are provided by separate manufacturers and with separate network infrastructures and so there is not much integration and interaction among different subsystems. In this paper, we envision a new energy sensing and monitoring framework with integrated communication backbone in the intelligent building environments. Specifically, through comprehensive comparisons and investigations, we study different candidate communicating media and protocols like wireline, wireless, and power-line communications technologies that potentially can be used in the intelligent buildings to realize the goals of coordination, integration, and energy efficiency. Also, we propose an extension "smart box" for integration of the devices before the maturity of the standardization process. Cloud computing and smart phone technologies are also introduced to realize the goals of improving energy efficiency and promote global sustainability.

Abstract: A variable rate fertilizer application system with ARM microprocessor as control core was developed. The structure and realization principle of the system were introduced. The system consists of two velocity sensors of photoelectric encoder and proximity transducer, an ARM controller, and an executive unit of stepper motor. The system has automatic and manual variable rate fertilization modes. In the automatic mode, the system can be positioned automatically without GPS, and controls the stepping motor’s rotational speed according to the fertilizing amount in different grid to realize variable rate fertilization. In the manual mode, the controller combines the fertilizing amount in different grid input manually through the keys with the velocity of the applicator, and calculates the stepping motor’s rotational speed to realize the variable rate fertilizer application. The applicator working status can be seen in LCD on the controller. The experimental results show that the system can work properly.

Abstract: An embedded controller in green house is designed and analyzed in this paper, which is consisted of embedded terminal and wireless sensor node. The captured data by sensor nodes will be sent to the base station, after which, certain information will be sent to controller by the base station through RS232 protocol. The controller is used for receiving, displaying, storing the temperature information and controlling the temperature according to some related settings, which is implemented with ARM 9 micro processor and embedded Linux OS. The system is confirmed to be stable and reliable in field practice, which fulfills the design specification

Abstract: According to the problem that manual controlling process for greenhouse is complicated and low accuracy, this paper designed and realized an intelligent temperature and humidity monitoring system for greenhouse based on ARM11 embedded device and WinCE operating system. This system consisted of wireless sensor, embedded device, drive circuit and execution units. It collected the temperature and humidity of the greenhouse, and then drove the fan, mechanical shutter and sprinkler to regulate the temperature and the humidity. The system is simple, high reliability, and helpful to reduce farmers’ labor and increase efficiency.