Industry 4.0 production comprises complicated highly automated processes. However, human activities are also a crucial component of these processes, e.g., for machine main- tenance. Task assignment of human resources in this domain is challenging, as many factors have to be taken into account to ensure effective and efficient activity execution and satisfy special conditions (like worker safety). To overcome the limita- tions of current Business Process Management (BPM) Systems regarding activity resource assignment, this contribution provides a BPM-integrated approach that applies fuzzy sets for activity assignment. Our findings suggest that this approach can be easily applied to complex production scenarios, while providing efficient performance even with a large number of concurrent activity assignment requests. Additionally, our evaluation shows its potential for improved work distribution which can lead to cost savings in Industry 4.0 production processes.
Human emotion prediction is an important aspect of conversational interactions in social robotics. Conversational interactions involve a combination of dialogs, facial expression, speech modulation, pose analysis, head gestures, and hand gestures in varying lighting conditions and noisy environment involving multi-party interaction. Head motions during conversational gestures, multi-agent conversations and varying lighting conditions cause occlusion of the facial feature-points. Popular Convolution Neural Network (CNN) based predictions of facial expressions degrade significantly due to occluded feature-points during extreme head-movements during conversational gestures and multi-agent interaction in realworld scenarios. In this research, facial symmetry is exploited to reduce the loss of discriminatory feature-point information during conversational head rotations. CNN-based model is augmented with a new rotation invariant symmetry-based geometric modeling. The proposed geometric model corresponds to Facial Action Units (FAU) for facial expressions. Experimental data show hybrid model comprising a CNN-based model and the proposed geometric model outperforms the CNNbased model by 8%-20%, depending upon the type of facialexpression, beyond partial head rotations. Keywords-Artificial Intelligence; conversation; emotion analysis; facial expression analysis; facial occlusion; facial symmetry; head movement; multimedia.
The present paper aims to control selective information to understand the main mechanism of information processing in multi-layered neural networks. We propose two types of selective information, namely, individual and collective selective information, or simply, individual and collective information. The individual information represents to what degree a neuron is connected specifically to another one, and it should be increased as much as possible. Then, we try to use this abundant information as impartially as possible, reducing the specificity of collective neurons and reducing collective information. By controlling the ratio of individual and collective information, we can realize a number of different types of states to be interpreted, leading to the interpretation of the inference mechanism. The method was applied to the bankruptcy data set. In the experiments, we successfully increased individual information and decreased collective information. By examining partially compressed weights, we could see how neural networks, by controlling the selective information, can process information content in multi-layered neural networks. This examination of information flow can lead us to understand the main inference mechanism of neural networks. Keywords—individual, collective, information, selectivity, partial compression, interpretation, generalization
—Enabling humans and robots to work together allows for cost savings and workplace efficiency. The robot must be equipped to perceive humans, and redirect its actions for co- operative tasks or under hazardous situations. Thus, dynamic motion planning appears as an essential exercise. We use dynamic roadmaps for online motion planning in changing environments combined with a voxel-based grid. The presented approach can answer path planning queries efficiently. Visualized simulation is an important technique for rapid verification of algorithms or prototypes. We present an architecture that implements a simulation of the robotic manipulator using the Robot Operating System (ROS) and MoveIt.
Interacting with the environment, many robots would benefit from advanced tactile sensors complementing optical sensors in particular when operating under poor visibility. In nature, rats exhibit a prominent tactile sense organ, the socalled vibrissae. For instance, these enable rats to detect shape and texture information of objects based on few contacts. Since vibrissae consist of dead tissue, all sensing is performed in the support of each vibrissa, the follicle-sinus complex. Inspired by this characteristic measuring principle, we set up a mechanical model, consisting of a cylindrical, one-sided clamped bending rod, which is swept along a 3D object surface undergoing large deflections. In doing so, the focus is on both simulating the scanning sweep in order to determine the support reactions of the rod during object scanning and subsequently using these quantities in order to reconstruct a sequence of contact points as a basis for shape reconstruction. The simulated scanning sweeps include tip and tangential contacts, as well as longitudinal, lateral and axial slip. The object reconstruction reveals that simple scanning kinematics, e.g., passive dragging of the tactile sensor on a mobile robot, are sufficient in order to capture fragments of object shapes and thus to complement data gathered by optical sensors. Keywords–Vibrissa; tactile sensor; surface sensing; surface reconstruction;
To ensure the safety and security of workplaces and homes, parts replacement of mechanical/electric/electronic devices is inevitable, because they will gradually break down. Users have to search for the appropriate part numbers from a parts database to order and replace the parts. In many cases, users make phone calls to the support center of the vendors instead of searching by themselves, which can be inconvenient for the support center staff. In this paper, to reduce the number of phone calls to support centers, we propose a prototype of an automatic part number answering system from natural language. This system consists of open-source speechto-text conversion and Structured Query Language (SQL) generation from the natural language. Preliminary evaluation results show that 83% of the voice questions returned the correct part numbers. In addition, the search with our system was executed an average of 3.86 times faster than with the conventional manual keyword search. Keywords-speech-to-text; SQL generation; natural language processing.
The latest advances in industry have been boosted by the application of the Industrial Internet of Things (IIoT), which is being supported by the implementation of Cyber-Physical Production Systems (CPPS). In this context, simulation and optimization models have, for some time now, been used to reduce CPPS design complexity, implementation time, and operating costs throughout its life cycle. Considering the complexity and heterogeneity of CPPS components and their different application areas within the manufacturing context, simulation models may contain unreliable representation of the real system, since the behavior of the CPPS in a physical environment can be quite different from the same CPPS, specified in a simulation model, considering the same events. Thus, facing this inherent difficulty in reliably modeling a CPPS in a virtual environment, the authors propose a tool Layout CAD Interface, which enables simulation modeling software with automatic generation capabilities of discrete-event simulation models. The tackled simulation software was Simio, and the generation of simulation models was based on Computer-Aided Design (CAD) files referring to shop-floor layouts. Keywords–Cyber-Physical Production System; Simulation; Simio; CAD.
Because the principle of a certain type of tank fire control system is complex and the control signal is transient, the fire control system of a certain type of tank mainly depends on the distribution of detection equipment for fault detection, but there is no corresponding rapid detection equipment for the maintenance personnel in the process of its work, resulting in a single detection means. Troubleshooting is difficult. By studying the core component of gun control system, the detection port of gun control box. Measuring the signal of the detection port of the gun control box is the basic method to eliminate the fault of the fire control system. It is very necessary to measure the signal of the detection port quickly and conveniently. Therefore, the demand of rapid detection equipment for gun control box is obvious. This paper designs and manufactures the rapid detection equipment of the detection port, which provides the corresponding technical means for the rapid detection and fault diagnosis of a certain tank fire control system in the army. The signal value of each pin of the socket of the test port of the gun control box is to show the status value of the normal operation of the fire control system, and to detect the signal of each pin can quickly isolate the components of the system fault, and determine the fault location by signal fault tree analysis, so as to improve the speed of judging the fault. The test equipment of gun control box is mainly used to measure the signal of the test port of gun control box. When fault detection is carried out, the signal to be measured can be obtained quickly and conveniently for analysis. Therefore, the starting point of designing and manufacturing the rapid detection equipment for the gun control box is mainly to provide a simple and convenient detection equipment for troubleshooting a certain type of tank fire control system, and to provide data information for fault analysis. 1. The Design Principle of the Detection Box A certain type of tank gun control box rapid detection equipment, mainly is the gun control box detection port signal acquisition input to the multi-channel conversion circuit, through the single-chip computer keyboard management design, control keyboard input to select the signal to be measured, analog channel switching circuit design, can be the gun control box detection signal of any signal value on-off, in the display Dynamic display of data and signal waveform on the oscilloscope (using a small "digital oscilloscope", digital oscilloscope can not only display the size of the signal, but also display the waveform and transient state of the signal. Corresponding detection signals can be quickly obtained, which is convenient for quantitative analysis of the system, and then to determine the failure parts. The block diagram is shown in the diagram[1] . When the function of the detection box is expanded, the detection ports of other components can be connected. Just connect the pin number of the port according to the programming address, the pin number and keyboard program can be used for other port measurement. Of course, the number of ports connected is limited[2] . 2. The Realization Principle of the Detection Box Detection box uses the management and control function of single chip microcomputer to realize selective and purposeful detection of signal pins to be measured. The signal pins of gun control box 2019 2nd International Conference on Intelligent Systems Research and Mechatronics Engineering (ISRME 2019) Copyright © (2019) Francis Academic Press, UK DOI: 10.25236/isrme.2019.010 49 are all connected to the detection box by means of switching connectors. When the signal value of a pin needs to be measured, the corresponding number is input on the keyboard and controlled by single chip microcomputer to be connected to the detection box. The signal circuit in the measuring box is connected with the oscillograph and the magnitude of the measurement is displayed on the oscilloscope. The numeric value of keyboard input is displayed by digital tube, which is convenient for accurate input. 3. Circuit and Software Design Hardware design principle: the detection box mainly consists of five parts: gun control box signal acquisition interface, multi-switch conversion circuit, keyboard input and read, single-chip microcomputer using Atmel AT89C52 core processor, micro-digital oscilloscope temporary selection model, the system uses 11.090MHz crystal oscillator, two digital tubes, single-chip microcomputer for the P0 port. At the input control end of the keyboard, the lower four bits are rows, the higher four bits are columns, the P1 port is used for the display of the digital tube, the P2 port is the output port, the P2.0-P2.3 is the segment selection, and the P2.4-P2.7 is the bit selection input terminal[4] . The technical route of the scheme is to collect signal by interface line, input data by keyboard, control signal conduction by AT89C52, display signal value by micro-digital oscilloscope, read by 4 ×4 matrix keyboard, draw schematic diagram by Protel software, and complete software programming by keil software. When the function of the detection box is expanded, the detection ports of other components can be connected. Just connect the pin number of the port according to the programming address, the pin number and keyboard program can be used for other port measurement. Of course, the number of ports connected is limited. 3.1. Core Processor AT89C52. 89C52 is the basic product of the MCS-51 series microcontroller of INTEL company. It adopts the reliable CMOS technology of Atmel company to manufacture the high-performance 8-bit microcontroller. It belongs to the standard MCS-51 HCMOS product. It combines the high-speed and high-density technology of CMOS with the low-power characteristics of CMOS. It is based on the standard MCS-51 microcontroller, architecture, and Instruction system. It is an enhanced version of 89C51 MCU, which integrates clock output and up or down technology, and is suitable for various control applications. 89C52NEIZHI 8-bit CPU, 256-byte internal data memory RAM, 8K on-chip program memory (ROM), 32 bidirectional input/output (I/O) ports, 3 16-bit timing/counter and 5 two-stage interrupt structures, a full-duplex serial communication port, on-chip clock oscillation circuit. In addition, 89C52 can also work in low power mode, and can select idle and power down mode through two kinds of software. Freeze CPU in idle mode, while RAM timer, serial port and interrupt system maintain its function. Power down mode, to maintain RAM data, always stop and stop other functions in the chip. 89C52 has PDIP (40 product pin) and PLCC (44pin) two package forms. Structural features: 8-bit CPU; on-chip oscillator and clock circuit; 32 with I/O interface; external memory addressing range ROM, RAM64K; 2 16-bit timing/counter; 5 interrupt sources, 2 interrupt priority; full duplex serial port; Boolean processor. 3.2. 4×4 Matrix Keyboard. The parallel port P0 of the MCS-52 is connected with a 4*4 matrix keyboard. The low 4-bit P 0-P 0.3 is used as the row and the high 4-bit P 0.4-P 0.7 is used as the row. Matrix keyboard can effectively save chip port and high efficiency. Matrix keyboard is shown in Fig.1.
This paper introduces the importance of soil moisture to plant growth, the composition of soil moisture and the expression of soil moisture. Soil moisture measurement techniques are classified according to sampling method. The principle and test method of six typical soil moisture measurement methods, including drying method, tensile method, neutron emission method, infrared ray method, dielectric method and dew point microvoltmeter method, were analyzed in detail, and their advantages and disadvantages were summarized. It can be used for reference in the future research, application and deep multi-disciplinary integration of soil testing technology.
In this paper, we demonstrate how to analyze the WiFi data of the German highspeed trains called InterCityExpress (ICE) on the basis of a neural network. To achieve this, we apply a Self-Enforcing Network with cue validity factors to underline the importance of selected features. It is shown that the quality of the WiFi connection, in terms of the rate of downloads and the latency, can be grouped and explained by just a few determinants. We will show where the network coverage is especially good or bad and suggest ways to improve this quality to enhance the comfort of traveling on the highspeed trains and therefore to possible expand the profits of the operating company. Keywords–Self-Enforcing Network (SEN); self-organized learning; cue validity factor; Intelligent data analysis; Industry 4.0 data analysis.
Abstract: This paper studied the dengue fever model with time delay. This paper divided the time delay into four cases: (1) τ1 = τ2 = 0, (2) τ1 = τ, τ2 = 0, (3) τ1 = 0, τ2 = τ , (4) τ1 = τ, τ2 = τ , and studied the stability and Hopf bifurcation of the model on these three cases. At the end of this paper, we simulated the dengue model with time delay by using Matlab software, and gained the numerical condition of this model which appearing periodic solutions and Hopf bifurcation. On the first case τ1 = τ, τ2 = 0, the time delay threshold is τ0 = 0.6155; on the second case τ1 = 0, τ2 = τ , the time delay threshold is τ0 = 0.0490; on the third case τ1 = τ, τ2 = τ , the time delay threshold is τ0 = 3.5454.
Periodic pattern mining in time series is of great practical significance to scientific research and practical application as it is beneficial to improve prediction and trend analysis. Current algorithms for periodic pattern mining in time series can be divided into three categories according to whether the parameters of periodic length are known or not. This paper discusses the specific classification of periodic patterns and summarizes advantages and disadvantages of these algorithms, with a purpose to provide reference for the practical application of these algorithms.
It is one of the important tasks for human data acquisition to extract the human object with the specific action. However, the accuracy of object extraction will be influenced by the shadow in an image as well as the non-standard human action. To address this issue, we propose a novel human object extraction method. First, we extract the moving object contour with the combination of three-frame-difference method and mixture gaussian model. Next, we remove the shadows from the image with multi-feature fusion method. Then, we extract the human skeleton with distance transformation. Finally, we select the image according to the angle of skeleton. The experimental results show that the proposed method can accurately extract the human object with the specific action from the image.
This paper analyzes the proportion of different charging power levels by traversing optimization algorithm, which can meet the needs of users, reduce equipment investment and reduce the peak-to-valley difference of electric vehicle charging load. First, the minimum charging power Pmin of the electric vehicle is obtained by using the charging power in the known data than the charging time, and the overall data is sorted according to Pmin, and the charging power levels of the sorted electric vehicles are ranked by 10%, 40%, and 50%. Divided into three levels, the actual charging time of the electric vehicle is obtained according to the charging power of the electric vehicle. This value is added to the starting charging time to obtain the stopping moment of the electric energy flow, and the total sum of the powers of all the vehicles in each time period is accumulated to obtain the sum. Charging load curve. For solving the optimal charging power level ratio, first set the ratio of the three levels to X, Y, Z, and then use the same method as the 24-hour charging load to establish the constraint condition that meets the user's demand with the unknown number, and then the device. Investment Ymin and peak-to-valley difference P ∆ respectively give a weight of 1:1 to establish an Amin objective function, traverse all possible proportions, and find that when the ratio of three levels is 48:51:1, the minimum equipment investment is 1409 million yuan, the smallest peak valley The difference is 18810kW.
This paper firstly we established an integer programming optimization model according to the type and quantity of medical packages required by each hospital, and obtained the type of the drone cargo bay required to transport the medical packages, and then determined the types of drones (C, E, F, G) that transport the medical packages. Combined with the actual situation of drone flight range, the option to use one or two standard containers was excluded. Finally, we established the DroneGo disaster response system with three ISO cargo containers drone to enhance Puerto Rico's ability to respond to hurricane disasters.
This paper uses the first principle of Density functional theory (DFT) to theoretically calculate the electronic structure parameters and optical properties of MgB2 superconductors. In the analysis of electronic structural parameters, the unitary model of MgB2 was optimized by generalized gradient approximation and local density functional approximation, and the lattice parameters of the model were adjusted. The dielectric constant spectrum of photon energy and MgB2 , the refractive index of MgB2 , the photon reflection spectrum, the photon absorption spectrum and the energy loss function of MgB2 were obtained by calculation. Through the comparison of various relational graphs of MgB2, the conclusion that they have a mathematically proportional correspondence is obtained.
