Research on an Intelligent Diagnostic Equipment Suitable for Rail Traffic Braking System

In order to meet the increasing requirements of train safety in the rail transit industry, it is important for the operators to monitor the running status and health status of the train in real time. Based on train on-broad data, using big data analysis method for data mining to study the health status of trains and change rules, it will be helpful to realize the informationization and intelligence of train operation. This paper designs and introduces a smart device for the brake system of rail transit trains. It reads the required data through the train CAN Bus network, and uses the wireless transmission device DTU to remotely send the analysis result to the cloud server. The user can easily browse the train’s device status and system health analyzed by the server through the webpage anytime and anywhere, which is helpful for the health operations and management of trains. The equipment passed the relevant tests of the Changsha subway train in China, which can meet the needs of functional design.

Pilot-Induced-Oscillations Issue in Driverless Car Age

As driverless vehicles becoming more and more popular due to the development of artificial intelligence, human beings will gradually get free from the vehicle driving. However, unexpected oscillations may happen due to the unfamiliarity of the vehicle configuration when humans want to drive themselves, even the vehicle itself is stable. These driver-induced-oscillations are similar with the pilot-induced-oscillations (PIO) which is generally related with actuator rate limit in the aircraft systems. Thus, this study attempts to review the PIO issue briefly and provide a guidance to solve the potential human-in-the-loop unmanned driving challenge associated with rate limit effect.

Study of Industry 4.0 and its Impact on Lean Transformation in Aerospace Manufacturing

Lean manufacturing practices focus on minimizing all forms of waste from the production system. The applicability of lean manufacturing concepts and principles has often been questioned in sectors including aerospace manufacturing primarily due to their high variety - low volume environments. The key challenges include the difficulty in changing the factory layout, lack of plant-specific manufacturing strategies, lack of benchmarking between manufacturing plants and non-existence of learning through experimentation culture. Consequently, the aerospace manufacturing industries have struggled to implement lean principles over the years successfully. Industry 4.0 is a new paradigm that is significantly influencing several manufacturing industries across the globe. Applying the concepts of Industry 4.0 along with the conventional lean transformation technology could potentially address these challenges. The focus of this research is to study the possibilities of integrating Industry 4.0 tools with existing lean manufacturing philosophies within the aerospace manufacturing sector in order to improve various aspects of manufacturing processes in a cyber-physical environment. A case study is performed considering a quality inspection department in a typical aerospace industry. The case study is simulated using discrete event simulation tool — Arena. The study found that a hybrid approach involving the holistic merger of the lean principles along with the Industry 4.0 tools known as Lean Industry 4.0 is the best way forward for the aerospace manufacturing sector. The outcomes of this research provide an understanding of the role of industry 4.0 paradigm and their implementation in several other high-technology and high-risk manufacturing sectors including life sciences, space, and defense industries.

Model Predictive Control of General Fractional Order Systems Using a General Purpose Optimal Control Problem Solver: RIOTS

RIOTS is a Matlab toolbox capable of solving a very general form of integer order optimal control problems. In this paper, we present an approach for implementing Model Predictive Control (MPC) to control a general form of fractional order systems using RIOTS toolbox. This approach is based on time-response-invariant approximation of fractional order system with an integer order model to be used as the internal model in MPC. The implementation of this approach is demonstrated to control a coupled MIMO commensurate fractional order model. Moreover, the performance and its application process is compared to examples reported in the literature.

Electrode-Skin Impedance Component Estimation in the Time-Domain

This paper presents a method to estimate the individual component values of a bipolar electrode-skin interface, with future intent of applying compensatory electrode-skin impedance balancing prior recording bio-signals with electromyography. The electrode-skin interface was stimulated by a step input and the output behaviour was characterised using a single exponential model per electrode. The method was applied to simulated circuitry, passive component circuitry and a human subject. The accuracy of the method was determined using the known values that comprised the simulated and passive component circuitry. Nine of ten simulated data sets resulted in accurate estimations, with a maximum error of 0.763% and a mean error of 0.076% per component. The method also produced successful estimates for nine of the ten physical circuitry data sets, with a maximum error of 10.2% and a mean error of 3.49% per component. The method was unsuccessful in estimating the individual electrode-skin impedance components for the human subject: this was due to the system failing to reach steady state during the stimulation period. The authors suspect a DC offset caused by the half-cell potentials associated with the electrode-skin interface were the cause of the unexpected behaviour.

Development of an Assist-As-Need Controller for an Upper-Limb Exoskeleton With Voluntary Torque Estimate

This study considers the development of an assist-as-need torque controller for an exoskeleton for stroke rehabilitation. Studies have shown that active patient participation improves the patient’s recovery from stroke. Assist-as-need control, providing the patient with the assistance they need to complete a task, is desirable, as the assistance can be varied to maximise patient participation. However, research is limited, and current methods cannot guarantee optimal assistance as non-zero assistive forces are still provided to subjects that are capable of completing the task unassisted. This study proposes a control system to vary and optimise the assistance for a subject based on their capabilities. A particle filter developed from previous research is used to estimate the subject’s voluntary effort. The assistive torque is determined from a target torque and the voluntary effort. The controller is shown to be effective, as zero assistance is provided to a subject capable of completing the task unassisted. Additionally, the assistance will increase if the subject fatigues. Using the estimate of the subject’s strength, the assistive torque can be accurately set to maximise a patient’s participation, and therefore, the assist-as-need controller can lead to improved therapeutic results.

Renormalization Group Method for Singular Perturbed Systems Driven by Fractional Brownian Motion

In this article, we use the renormalization group method to study the approximate solution of stochastic differential equations (SDEs) driven by fractional Brownian motion with Hurst parameter H∈12,1. We derive a related reduced system, which we use to construct the separate scale approximation solutions. It is shown that the approximate solutions remain valid with high probability on large time scales. We also expect that our general approach can be applied to the fields of physics, finance, and engineering, etc.

Embedded RIOTS: Model Predictive Control Towards Edge

RIOTS is a general purpose optimal problem solver written as a MATLAB toolbox with mixed-language programming (C, Fortran, Matlab, Simulink). This first paper introduces how to make RIOTS run under an embedded platform RP3B (Raspberry Pi 3 B) with Windows 10. We presented the system architecture and a complete demo on running RIOTS as the inner kernel for MPC, using a house made thermal control system based on Peltier modules.

Affine ICP for Fine Localization of Smart-AGVs in Smart Factories

With the emergence of the concept of Industry 4.0, smart factories have started to be planned in which the production paradigm will change. Automated Guided Vehicles, abbreviated as AGV, that will perform load carrying and similar tasks in smart factories, Smart-AGVs, will try to reach their destinations on their own route instead of predetermined routes like in today’s factories. Moreover, since they will not reach their targets in a single way, they have to dock a target with their fine localization algorithms. In this paper, an affine Iterative Closest Point, abbreviated as ICP, based fine localization method is proposed, and applied on Smart-AGV docking problem in smart factories. ICP is a point set registration method but it is also used for localization applications due to its high precision. Affine ICP is an ICP variant which finds affine transformation between two point sets. In general, the objective function of ICP is constructed based on least square metric. In this study, we use affine ICP with correntropy metric. Correntropy is a similarity measure between two random variables, and affine ICP with correntropy tries to maximize the similarity between two point sets. Affine ICP has never been utilized in fine localization problem. We make an update on affine ICP by means of polar decomposition to reach transformation between two point sets in terms of rotation matrix and translation vector. The performance of the algorithm proposed is validated in simulation and the efficiency of it is demonstrated on MATLAB by comparing with the docking performance of the traditional ICP.

Analysis of Octree Coding for 3D Point Cloud Frame

The octree coding has been used extensively for 3D point cloud compression. For encoding, the recursive octree partition is used for the 3D point cloud frame until only one point is contained in the cube bin, and the representative bits are recorded. For decoding, the reversal procedure is performed to reconstruct the 3D point cloud frame. During the reconstruction, the reconstructed point in a particular bin is usually quantized to a particular corner of the bin for the reconstructed 3D point cloud frame; this introduces coding distortion in the point location. In this paper, nine reconstructed locations are tested to distinguish their distortion results, and the bit cost is also analyzed.