The deadbeat control law was used to control the constant current output of an inverter welding power source based on a full-bridge power converter topology. The controller has good dynamics; however, it is sensitive to changes of the inductance in the output welding loop and is not stable. The cable in the output welding loop of the power source provides an extra parasitic inductance, which leads to instabilities of the deadbeat control system. The pole assignment method was used to improve the deadbeat control and the control system adaptability to inductance variations. Tests show that the control system stability is enhanced by increasing the number of poles as well as the distance between the poles and the origin, but these changes slow the system dynamic response. The compromise between the stability and the response speed is balanced by adjusting the pole locations and the number of poles to provide the desired control characteristics of the welding power source.

The performance of numerical control (NC) rotary tables degrades slowly and oscillatorily. A Wiener process model is developed to improve the reliability estimate of NC rotary tables using the maximum likelihood method to analyze the measured characteristics of each parameter with fixed drift and diffusion coefficients assuming that the initial performance degradation values obey a normal distribution. The model gives the correlation reliability function for the NC rotary table. Comparison of the results with those of the pseudo life distribution method shows that the Wiener process model more accurately predicts the actual characteristics.

Austenitic thick-wall welds are difficult to evaluate using ultrasonic testing because the ultrasonic waves are scattered and attenuated by the coarse grain microstructure. The ultrasonic signals then contain much noise with signal processing needed to reduce the noise. This paper presents an ultrasonic signal extraction algorithm based on the Gaussian modulated pulse (GMP) model to improve the signal processing. The original signals are filtered by a cross-correlation method and then the filtered signals are decomposed to acquire the control parameters for each component based on the GMP model and the signal features. Simulations and experiments verify the algorithm's performance. This algorithm can extract useful information about defects from ultrasonic signals as the foundation for further analyses of thick-wall welds for defect information analyses and inspection method optimization.

The relationship between the spray process parameters and the atomizer spray cone angle must be well understood for establishing accurate spray gun models. However, traditional experimental methods cannot easily provide sufficient experimental data, such as measurements of the spray pattern size. This paper presents a method to measure the atomizer cone angle by using image processing to identify the edges on an atomizer cone image. This method was then used in an experimental study of the relationship between the shaping air pressure, the atomizing air pressure and the flow rate on the atomizer cone angle. The results show that the shaping air pressure is positively correlated with the angle of the long axis of the atomizer cone and negatively correlated with the angle of the short axis of the atomizer cone. The atomizing air pressure is positively correlated with the angles of the long and short axes of the atomizer cone and the fluid flow rate. The atomizing air pressure is also coupled with the shaping air pressure. The fluid flow rate has a relatively small effect on the atomizer cone angle.

Cargo compartment fire has become the major security threat for cruising aircraft that can cause huge property and casualties. The effects of low pressures on common solid fuel fire behavior in air freight was studied using different amounts of corrugated cartons can represent typical Class A solid fuel fires. Mass burning rate, radiative heat flux, and flame axial temperature had been measured as the principal characteristic parameters during the combustion process. The modified B-number theory was employed to model the burning rate behavior of solid fuel. The theoretical relationship between the burning rate and the ambient pressure for a turbulent solid fuel fire is ṁ"∝ C_hP^2/3+CrP^3/2. The experimental results showed that the modified pressure index is 1.3. In addition, the flame radiant heat flux and centerline temperature distribution were analyzed to indicate that the radiant heat flux first increased and then decreased. The radiant heat flux at the R2 location was greater than at other locations. The maximum flame temperatures was 800℃ in the flame zone. The solid fuel fire behavior under reduced pressure environment was investigated from both theoretical and experimental aspects, which provided a basis for aviation fire insurance.

Since communities are the basic units for public safety management, community risk prevention is of great significance. Community risk prevention must first identify community risks for people, things and management. This study analyzed the characteristics and causes of community risk to identify community risk prevention methods and how to monitor, control, predict, quickly detect and prevent community risk. Current international development trends for community risk prevention are reviewed to show that big data platforms are the key technology for community risk prevention. Finally, this paper describes the function, structure and construction of a large data platform for community risk prevention. This research on community risk prevention and big data platforms provides theoretical and technical support for community safety and security.

The impact of water droplets on a smooth stainless-steel surface was studied experimentally for various velocities and angles. The experiments studied the effects of impact angle and Weber number on the dynamic characteristics of the impact process, such as the displacement of leading and trailing edges, the maximum spread diameter and gliding length. The results showed that the maximum spread diameter increased with increasing Weber number, but decreased with increasing surface slope. However, once the water droplet reached equilibrium, the droplet contact diameter was found to be independent of the impact angle and the Weber number but mainly depended on the advancing and receding contact angles of the droplet on the surface. In addition, increasing both the impact angle and the Weber number increased the gliding length.

Robots need to identify the type of object, detect the shape and judge the distance to the object. This paper presents an object recognition and localization method that uses binocular information based on the Mask R-CNN model. The Mask R-CNN is used to process the binocular image and complete the bounding box selection, recognition and shape segmentation for each image. Then, the neural network feature is used to match the same object in the binocular images. Finally, the iterative closest point (ICP) method is used to estimate the parallax and calculate the distance according to the obtained object shape. Tests show that the method can process data in near real-time speed with better precision than the traditional disparity map algorithm.

Railway wireless data communications have problems with insufficient bandwidth and poor reliability. This paper describes a railway software-defined wireless communication network (RailSDN) which separates the control plane from the data plane, enables heterogeneous network interoperability among operators, provides user logical isolation using virtualization, and provides transparent data transmission for users. A network resource scheduling model is developed based on a Markov decision process which takes into account various factors such as the infrastructure network status and the user requirements. The system searches for the optimal link allocation with the optimization goal being to satisfy user demand and dynamically adapts to the user requirements for the network resources. Tests show that this railway wireless communication network is scalable, enables flexible customization of user demand, and improves the network resources utilization efficiency.

Many monitoring methods have been proposed for network measurements that are essential in software defined networking. However, periodically or adaptively collecting statistics from software switches using per-flow queries incurs significant communication costs thus increase the loads on switches. This paper presents an approach called OpenCost that decides which switch us used to collect statistics in software defined networks based on a non-linear integer programming (NLIP) model. However, the NLIP problem is NP-hard; therefore, the problem is solved using an approximation algorithm based on a greedy algorithm. Extensive simulations were used to benchmark the algorithm with the results showing that OpenCost reduces the communication costs by 55% on average compared with other methods.

Frequency hopping is an important anti-jamming and anti-interception technique that is widely used in communication systems. Frequency hopping is then combined with cryptographic algorithms to offer better security. This paper presents a frequency hopping communication system based on an optimized ZUC algorithm with three layers. The introduction of a permutation polynomial in the top layer increases the linear complexity while the evolutionary DES algorithm in the bottom layer achieves high security of the sequences. Security and randomness tests show that the resultant sequences have good hamming correlations, randomness, and linear complexity and pass the NIST test as well. Finally, the system was successfully implemented in the Simulink system.

In-wheel motor driven vehicles are powered directly by in-wheel motors installed inside the wheel which provide independent control of the driving and electrical braking torque. This paper presents a torque vectoring control system (TVCS) to improve the coupling and complex matching of the in-wheel motors. This system is independent of other systems in the vehicle and, thus, is a power assembly for electric vehicles. The paper describes the TVCS network topology and the software and hardware architecture of the torque vectoring controller. An in-wheel motor driven vehicle was built equipped with TVCS with the single-lane change test showing that the TVCS can effectively lower communication load rates and facilitate integration of the in-wheel motors. The test vehicle provides the basic system functions in a test platform for development and optimization of the torque vectoring control algorithm.