Hee-Hyol Lee

On-line tuning PID parameters in an idling engine based on a modified BP neural network by particle swarm optimization

PID control systems are widely used in many fields, and many methods to tune the parameters of PID controllers are known. When the characteristics of the object are changed, the traditional PID control should be adjusted by empirical knowledge. This may result in a worse performance by the system. In this article, a new method to tune PID parameters, called the back-propagation network modified by particle swarm optimization, is proposed. This algorithm combines conventional PID control with a back propagation neural network (BPNN) and particle swarm optimization (PSO). This method is demonstrated in the engine idling-speed control problem. The proposed method provides considerable performance benefits compared with a traditional controller in this simulation.

Real-time traffic signal learning control using BPNN based on predictions of the probabilistic distribution of standing vehicles

In this article, a new method to predict the probabilistic distribution of a traffic jam at crossroads and a traffic signal learning control system are proposed. First, a dynamic Bayesian network is used to build a forecasting model to predict the probabilistic distribution of vehicles in a traffic jam during each period of the traffic signals. An adjusting algorithm for traffic signal control is applied to maintain the probability of a lower limit and a ceiling of standing vehicles to get the desired probabilistic distribution of standing vehicles. In order to achieve real-time control, a learning control system based on a back-propagation neural network is used. Finally, the effectiveness of the new traffic signal control system using actual traffic data will be shown.

Cooperative behavior control of robot group using stress antibody allotment reward

Lately, development in robotics for utilizing in both industry and home is in much progress. In this research, a group of robots is made to handle relatively complicated tasks. Cooperative action among robots is one of the research areas in robotics that is progressing remarkably well. Reinforcement learning is known as a common approach in robotics for deploying acquisition of action under dynamic environment. However, until recently, reinforcement learning is only applied to one agent problem. In multi-agent environment where plural robots exist, it was difficult to differentiate between learning of achievement of task and learning of performing cooperative action. This paper introduces a method of implementing reinforcement learning to induce cooperation among a group of robots where its task is to transport luggage of various weights to a destination. The general Q-learning method is used as a learning algorithm. Also, the switching of learning mode is proposed for reduction of learning time and learning area. Finally, grid world simulation is carried out to evaluate the proposed methods.

An Intelligent Context-Aware Learning System Based on Mobile Augmented Reality

Learning content using context-aware mobile technology, whether the content is manual or interactive, is hardly expected to arouse learners to interest or immersion because most of real-life environment is discrete from mobile content. For this reason, Augmented Reality is used to fix the drawback and to provide learners with an educational environment fit for desirable practice of the theory of situated learning. Increasing interest in Augmented Reality in recent years has led to multiple study efforts to build applications based on Augmented Reality, most of which require additional hardware or software, resulting in difficulties in establishing proper learning environment in the field of education. Therefore, in this paper we propose an intelligent context-aware learning system based on mobile Augmented Reality that provides a hassle-free desirable learning environment requiring nothing but a common mobile device.

Stochastic model of production and inventory control using dynamic bayesian network

Bayesian Network is a stochastic model, which shows the qualitative dependence between two or more random variables by the graph structure, and indicates the quantitative relations between individual variables by the conditional probability. This paper deals with the production and inventory control using the dynamic Bayesian network. The probabilistic values of the amount of delivered goods and the production quantities are changed in the real environment, and then the total stock is also changed randomly. The probabilistic distribution of the total stock is calculated through the propagation of the probability on the Bayesian network. Moreover, an adjusting rule of the production quantities to maintain the probability of the lower bound and the upper bound of the total stock to certain values is shown.

Neuro PID control of power generation using a low temperature gap

Power generation using a low temperature gap converts heat energy into electricity by using the temperature difference. In this article, a simulation model for power generation using a low temperature gap, which uses a circulation cycle with ammonia as the working fluid, is constructed as a linear multiple input/multiple output (MIMO) model which has 2 inputs and 2 outputs based on the step response method. A PID controller using a back propagation neural network is designed so that the difference in pressure between the turbine inlet and outlet is kept at 0.3 Mpa.

A learning control of unused energy power generation

In recent years, the development of new clean energy without dependence on fossil fuel has become urgent. This article proposes a learning control system for power generation using a low-temperature gap which has been designed to maintain the speed of a steam turbine in a real environment. This system includes nonlinearity and the characteristics of changing parameters with age and deterioration, as in the real environment. The evaporator, condenser, and turbine systems have been modeled, and a PID control with the ability to learn, based on a BackPropagation neural network, has been designed.

Traffic signal control based on a predicted traffic jam distribution

In this article, we propose a new method of traffic signal control based on the predicted distribution of traffic jams. First, we built a forecasting model to predict the probability distribution of vehicles being in a traffic jam during each period of the traffic signals. A dynamic Bayesian network was used as the forecasting model, and this predicted the probability distribution of the number of standing vehicles in a traffic jam. According to calculations by the dynamic Bayesian network, a prediction of the probability distribution of the number of standing vehicles at each time will be obtained, and a control rule to adjust the split and cycle of the signals to maintain the probability of a lower limit and a ceiling of standing vehicles is deduced. Through a simulation using the actual traffic data of a city, the effectiveness of our method is shown.

Parallel fuzzy inference on hypercube computer

In fuzzy database systems that have very large rules or fuzzy data, the inference time is much increased. Therefore, a high performance parallel fuzzy inference architecture is needed. In this paper, we propose a novel parallel fuzzy inference engine using Hypercube architecture. In this, fuzzy rules are distributed and executed simultaneously. The ONE-TO-ALL algorithm is used to broadcast the fuzzy input to the all nodes. The results of the MIN/MAX operations are transferred to the output processor by the ALL-TO-ONE algorithm. By parallel processing of the fuzzy rules, the parallel fuzzy inference algorithm extracts match parallelism and achieves a good speed factor. This architecture can be used in large expert systems or fuzzy database systems.

Real-time stochastic optimal control for traffic signals of multiple intersections

In this paper, a real-time stochastic optimal control method of traffic signal is modified. In addition, H-GA-PSO algorithm is proposed to search optimal traffic signals based on the stochastic model. The H-GA-PSO algorithm is a modified Hierarchical Particle Swarm Optimization (H-PSO) algorithm based on Genetic Algorithm (GA) processing. Finally, the effectiveness of the stochastic optimal control method with H-GA-PSO algorithm is shown through simulations at multiple intersections using a micro-traffic simulator.