Earticle

In order to overcome the limitations on human accidents and reduce experimental costs during autonomous driving, studies using virtual simulator is conducted. In addition, by using open source-based frameworks and middleware, we develop a module that can be applied immediately to a real vehicle using a model developed for autonomous driving that has been studied in a virtual environment. In this paper, we use autonomous driving environment using a game engine-based CARLA simulator, and propose a model predictive control module by interlocking open-source ROS (Robot Operating System) and Autoware with the simulator. In addition, ROS manages vehicle information and sensor data and transmits them to the Autoware. In order to control autonomous driving, the control end part is implemented using MPC (Model Predictive Control) among Autoware APIs, which provides necessary functions such as 3D map generation and NDT matching algorithm for autonomous driving, location identification, object recognition, and vehicle control.

In this study, we develop a time series based solar power failure determination algorithm that predicts its own generation amount by sharing power generation information with neighboring sites without relying on meteorological data of the Meteorological Agency and compares it with the power generation amount obtained in real time to determine the presence or absence of a failure. For the implementation of the algorithm, we design a prediction model based on deep learning using LSTM function and implement a model that predicts the amount of solar power that changes in real time. After the development of LSTM model, the RMSE was 93.85 as a result of preliminary test by comparing the predicted and the measured photovoltaic power generation. As the data learning process progresses and as the optimization process is continued, the prediction performance is expected to be further improved.

In this study, we develop a time series based solar power failure determination algorithm that predicts its own generation amount by sharing power generation information with neighboring sites without relying on meteorological data of the Meteorological Agency and compares it with the power generation amount obtained in real time to determine the presence or absence of a failure. For the implementation of the algorithm, we design a prediction model based on deep learning using LSTM function and implement a model that predicts the amount of solar power that changes in real time. After the development of LSTM model, the RMSE was 93.85 as a result of preliminary test by comparing the predicted and the measured photovoltaic power generation. As the data learning process progresses and as the optimization process is continued, the prediction performance is expected to be further improved.

이 연구는 도로 이용의 효율을 향상시키기 위하여, 거시적 시뮬레이션 기법인 Cell Transmission Model (CTM)과 유전 알고리즘(Genetic Algorithm)을 이용한, 클라우드 환경에서의 고속도로 교통 예측 및 최적 제어 시스템을 제안하였다. 본 연구에서 제안하고 있는 시스템은 4가지로 구성된다: 1) 전처리 모듈에서는 도로에 설치된 차량 검지기에서 수접된 데 이터를 1차적으로 처리하여 보정한다. 2) 데이터 기반 교통 예측 모듈에서는 계층적 구조 기반의 K-근접이웃 분류기법 (MK-NN)으로 1차 처리된 데이터를 분석하여 미래의 교통량과 스피드를 예측한다. 3) 온라인 교통 시뮬레이션에서는 MK-NN을 통하여 예측된 교통량과 스피드에 기초하여 사고, 공사, 날씨 등의 다양한 도로 환경에 따른 교통 상황의 변 화를 예측 한다. 4) 최적 교통 제어에서는 유전 알고리즘과 CTM을 기반으로 도로의 교통을 제어할 수 있는 최적 해법 을 도출한다. 본 연구의 분석에 따르면 제안된 알고리즘을 현재 고속도로의 교통 제어에 적용할 경우 평균 26%에서 66%의 총통행시간(VHT) 향상을 기대할 수 있다.

This study proposes the traffic prediction and optimal traffic control system based on cell transmission model and genetic algorithm in cloud environment. The proposed prediction and control system consists of four parts. 1) Data preprocessing module detects and imputes the corrupted data and missing data points. 2) Data－driven traffic prediction module predicts the future traffic state using Multi－level K－Nearest Neighbor (MK－NN) Algorithm with stored historical data in SQL database. 3) Online traffic simulation module simulates the future traffic state in various situations including accident, road work, and extreme weather condition with predicted traffic data by MK－NN. 4) Optimal road control module produces the control strategy for large road network with cell transmission model and genetic algorithm. The results show that proposed system can effectively reduce the Vehicle Hours Traveled upto 60%.

This paper presents methods of model predictive control (MPC) for eco-driving minimizing trip-time and energy- consumption of electric vehicles (EVs). Considering both non- convex and convex optimization problem formulations for MPC- based eco-driving, we compare the performances of nonlinear and linear MPC solutions for high-level planning of vehicle speed and charging in a driving simulation of a Munich– Cologne trip (573 km). The linear MPC can be considered as a convex quadratic programming approximation (i.e., convexified quadratic program) of the original nonlinear MPC, but its performance of optimality is shown to be comparable to the nonlinear counterparts whereas its computation speed is one order of magnitude faster.

This paper presents a model predictive control (MPC) based on a neural network (NN) model for air/fuel ration (AFR) control of automotive engines. The novelty of the paper is that the severe nonlinearity of the engine dynamics are modelled by a NN to a high precision, and adaptation of the NN model can cope with system uncertainty and time varying effects. A single dimensional optimization algorithm is used in the paper to speed up the optimization so that it can be implemented to the engine fast dynamics. Simulations on a widely used mean value engine model (MVEM) demonstrate effectiveness of the developed method.

To improve the accuracy of model predictive control, this paper presents an improved multiple cell linear parameter varying model predictive control method for carrier-based aircraft. After establishing the lateral dynamic model of carrier-based aircraft for multiple cell predictive controller, the output-feedback linear parameter varying control based on states observation should be implemented. The model simulation results indicate the better performance of the new method in comparison with the traditional controller with more accuracy and practicability.

This paper presents the method of multivariable integrated model predictive control for the nuclear power plant, including designing controller and building the modeling of turbine and once-through steam generator. The simulation results show that the rotate speed of turbine and outlet pressure of steam generator under the multivariable integrated model predictive control is faster steady and smaller overshoot than under the PID control, when the power load of nuclear power plant is changed. Simulation test indicates that the multivariable integrated model predictive control can obtain better control performance for nuclear power plant.

This paper addresses an optimal train trajectory planning and tracking problem for automatic train operation (ATO) with consideration of the train model uncertainty and constraints. Based on the discrete linear multi-points train model, an ATO control algorithm is presented to track piecewise reference by using model predictive control with constraints tightening such that the feasibility and robust convergence of this algorithm are guaranteed under the varying running resistance, automatic train protection (ATP) constraint, and train motor physical limits. Specifically, the features of the algorithm are: (i) taking traction and braking force of locomotives and braking force of carriges into account explicitly; (ii) integrating constraints tightening approach into piecewise reference tracking problem to ensure robustness; (iii) combining the optimal planning level and tracking control level together. Finally two case studies are conducted to verify the effectiveness of the algorithm.

Model predictive control is the family of controllers, makes the explicit use of model to obtain control signal. The reason for its popularity in industry and academia is its capability of operating without expert intervention for long periods. There are various control design methods based on model predictive control concepts. This paper provides review of the most commonly used methods that have been embedded in an industrial model predictive control. The most widely used strategies as Dynamic matrix control (DMC), Model algorithmic control (MAC), Predictive functional control (PFC), Extended prediction self-adaptive control (EPSAC), Extended horizon adaptive control(EHAC) and Generalized predictive control(GPC) have been described with history, basic idea, properties, and their controller formulation.

A model predictive control method is proposed for the point stabilization of wheeled mobile robots (WMRs) subject to nonholonomic constraints. The problem is simplified by considering only the steering system and neglecting the vehicle dynamics. A linearized error model is then formulated by transforming the robot position into polar frame. The feedback control policy is obtained by minimizing a quadratic cost function which penalizes the predicted errors and control variables in each sampling time over a finite horizon. The proposed control law is proven to guarantee the exponential stability of the robot system by considering additive inequality constraints in the optimization process. The performance of the stabilization algorithm is verified through computer simulations showing that the proposed method has a good regulation performance and convergence.

The modular multilevel converter (MMC) has been widely researched in the past decade for its superior performance in high voltage high power applications. However, applying MMC into electric motor drive systems is a intrinsically challenging problem because the capacitor voltages fluctuate in a wide range when the converter works at low output frequencies. A novel model predictive vector control (MPVC) scheme is presented to ensure the output phase currents’ quality of MMC for entire frequency range. In this scheme, model predictive control (MPC) is adopted to track the phase currents’ reference values and balance the capacitor voltages as far as possible. By replacing the conventional inner current control loops and pulse width modulation (PWM) module with MPC controllers, the vector control oriented by rotor flux is constructed to meet the requirements of drive systems. Simulations of a MMC drive with an induction motor are carried out using the Matlab/Simulink. The simulation waveforms demonstrate that the drive system achieves satisfied steady and dynamic performance.

Forecasting quality of complex products is a major concern for quality engineers and enterprises' decision-makers. But only a few researchers have investigated how multiple linear regression analysis can improve forecasts. This article presents a predictive control model for personalized forecasting system of complex product quality using multiple linear regression analysis. First, this paper compares the performance of personalized forecasting system for product quality which leads to a better understanding of the applicability of the forecasting model. Second, this paper identifies the effect of predictive control model of multiple linear regression analysis on forecasting accuracy. Then, an example shows the feasibility of using predictive control model in complex product quality personalized forecasting. Finally, it makes a conclusion that the approved prediction system embraces superiority over the prediction of the quality factors and manufacturing resources.

This paper proposes a fault-tolerant control strategy with finite control set model predictive control (FCS-MPC) based on hierarchical optimization for five-level H-bridge neutral-point-clamped (5L-HNPC) inverter-fed induction motor drives. Fault-tolerant operation is analyzed, and the fault-tolerant control algorithm is improved. Adopting FCS-MPC based on hierarchical optimization, where the voltage is used as the controlled objective, called model predictive voltage control (MPVC), the postfault controller is simplified as a two layer control. The first layer is the voltage jump limit, and the second layer is the voltage following control, which adopts the optimal control strategy to ensure the current following performance and uniqueness of the optimal solution. Finally, simulation and experimental results verify that 5L-HNPC inverter-fed induction motor drives have strong fault tolerant capability and that the FCS-MPVC based on hierarchical optimization is feasible.

This paper proposes a new method of Model Predictive Control (MPC) of nonlinear process by us-ing the measured Volterra kernels as the nonlinear model. A nonlinear dynamical process is usually de-scribed as Volterra kernel representation, In the authors' method, a pseudo-random M-sequence is ar plied to the nonlinear process, and its output is measured. Taking the crosscorrelation between the input and output, we obtain the Volterra kernels up to 3rd order which represent the nonlinear characteristics of the process. By using the measured Volterra kernels, we can construct the nonlinear model for MPC. In applying Model Predictive Control to a nonlinear process, the most important thing is, in general, what kind of nonlinear model should be used. The authors used the measured Volterra kernels of up to 3rd order as the process model. The authors have carried out computer simulations and compared the simulation results for the linear model, the nonlinear model up to 2nd Volterra kernel, and the nonlinear model up to 3rd order Vol-terra kernel. The results of computer simulation show that the use of Valterra kernels of up to 3rd order is most effective for Model Predictive Control of nonlinear dynamical processes.