• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2002.03a
• /
• pp.219-227
• /
• 2002

In this paper, we designed neural network predictive PID controller to control sway happened in transfer of trolley for automatic travel control system. We include dynamic character of nonlinear system, and mathematical expression veny simple used neural network. When various establishment location and surrounding disturbance were approved based on mathematical modelling of crane, controller designed to become effective control location error and vibration angle of two control variables that simultaneously can predictive control. Neural network predictive PID controller produced parameter of PID controller using neural network self-tuner. Neural network self-tuner's input used crane's output and neural network predictive output. Neural network self-tuner using error back propagation algorithm. We analyzed control performance comparison through computer simulation when applied disturbance about sway of location and angle in transfer of crane. The results show that the proposed neural network predictive PID controller has better performances than general PID controller, neural network PID controller.

• ETRI Journal
• /
• v.45 no.2
• /
• pp.318-328
• /
• 2023

Recently, embedded systems, such as mobile platforms, have multiple processing units that can operate in parallel, such as centralized processing units (CPUs) and neural processing units (NPUs). We can use deep-learning compilers to generate machine code optimized for these embedded systems from a deep neural network (DNN). However, the deep-learning compilers proposed so far generate codes that sequentially execute DNN operators on a single processing unit or parallel codes for graphic processing units (GPUs). In this study, we propose PartitionTuner, an operator scheduler for deep-learning compilers that supports multiple heterogeneous PUs including CPUs and NPUs. PartitionTuner can generate an operator-scheduling plan that uses all available PUs simultaneously to minimize overall DNN inference time. Operator scheduling is based on the analysis of DNN architecture and the performance profiles of individual and group operators measured on heterogeneous processing units. By the experiments for seven DNNs, PartitionTuner generates scheduling plans that perform 5.03% better than a static type-based operator-scheduling technique for SqueezeNet. In addition, PartitionTuner outperforms recent profiling-based operator-scheduling techniques for ResNet50, ResNet18, and SqueezeNet by 7.18%, 5.36%, and 2.73%, respectively.

• Proceedings of the Korea Society for Simulation Conference
• /
• 1999.10a
• /
• pp.247-253
• /
• 1999

In this paper predictive PID control system using neural network (NNPPID) is proposed to control temperature system. NNPPID is composed of neural network predictor forecasts the future output of plant based on the present input and output of plant. Neural self-tuner yields parameters of PID controller. Experiments prove that NNPPID temperature control system has better performance than conventional PID control.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.3
• /
• pp.322-327
• /
• 2003

The water level of a steam generator of pressurized light water nuclear Power generator is known as a subject whose control is difficult because of a shrinking and swelling effect that is been mutually contradictory in a variation of feed water. In this paper, a neural network model selects first coordinative controller by a inappropriate gain of two PI controllers and the selected controller＇s gain is tuned by a fuzzy self-tuner. Model inputs consist of the water level, the feed water, and the stream flow. One controller of both coupling controllers whose gain is handled firstly is decided based upon above data. The proposed method can analyze patterns of signals using the characteristic of neural networks and select one controller that needs to be tuned through the observed result in this paper. If one controller between both the water level controller and the feed water controller is selected by the neural network model then a gain of the PI controller is suitably tuned by the fuzzy self-tuner. Rules of the fuzzy self-tuner drew from the pattern of input and output data. In the summary, the goal of this Paper is to select the suitable controller and tune the control gain of the selected controller suitably through such two processes.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.346-349
• /
• 2003

For a traveling crane, various control methods such as neural network predictive control and TDOFPID(Two Degree of Freedom Proportional Integral Derivative) are studied. So in this paper, we proposed improved navigation method to reduce transfer time and sway with anti-collision path for avoiding collision in its movement to the finial coordinate. And we constructed the NNPPID(Neural Network Predictive PID) controller to control the precise move and speedy navigation. The proposed predictive control system is composed of the neural network predictor, TDOFPID controller, and neural network self-tuner. We analyzed ASC(Automated Stacking Crane) system and showed some computer simulations to prove excellence of the proposed controller than other conventional controllers.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2000.11a
• /
• pp.159-167
• /
• 2000

During the operation of crane system in the container yard, it is necessary to control the crane trolley position so that the swing of the hanging container is minimized. Recently an automatic control system with high speed and rapid transportation is required. Therefore, we designed a controller to control the crane system with disturbances and weight change. In this paper, we present the neural network two degree of freedom PID controller to control the swing motion and trolley position. Then we executed the computer simulation to verify the performance of the proposed controller and compared the performance of the neural network PID controller with our proposed controller in terms of the rope swing and the precision of position control. Computer simulation results show that the proposed controller has better performances than neural network PID with disturbances.

• Journal of Navigation and Port Research
• /
• v.26 no.5
• /
• pp.537-542
• /
• 2002

Recently, an automatic crane control system is required with high speed and rapid transportation. Therefore, when container is transferred from th intial coordinate to the finial coordinate, the container paths should be built in terms of the least time and no swing. So in this paper, we calculated the anti-collision path for avoiding collision in its movement to the finial coordinate. And we constructed the neural network predictive PID (NNPPID) controller to control the precise navigation. The proposed predictive control system is composed of the neural network predictor, PID controller, neural network self-tuner which yields parameters of PID. Analyzed crane system through simulation, and proved excellency of control performance than other conventional controllers.

• Proceedings of the KAIS Fall Conference
• /
• 2009.12a
• /
• pp.630-632
• /
• 2009

The parameters of PID controller should be readjusted whenever system character change. In spite of a rapid development of control theory, this work needs much time and effort of expert. In this paper, to resolve this defect, after the sample of parameters in the changeable limits of system character is obtained, these parametrs are used as desired values of back propagation learning algorithm, also neural network auto tuner for PID controller is proposed by determing the optimum structure of neural network. Simulation results demonstrate that auto-tuning proper to system character can work well.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.142-142
• /
• 2000

During the operation of crane system in container yard, it is necessary to control the crane trolley position so that the swing of the hanging container is minimized. Recently an automatic control system with high speed and rapid transportation is required. Therefore, we designed a controller to control the crane system with disturbances. In this paper, Ive present the neural network two degree of freedom PID controller to control the swing motion and trolley position. Then we executed the computer simulation to verify the performance of the proposed controller and compared the performance of the neural network PID controller with our proposed controller in terms of the rope swing and the precision of position control . Computer simulation results show that the proposed controller has better performances than neural network PID with disturbances.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.48 no.4
• /
• pp.434-443
• /
• 1999

In this paper, neural network predictive PID (NNPPID) control system is proposed to reduce the vibration of building structure. NNPPID control system is made up predictor, controller, and self-tuner to yield the parameters of controller. The neural networks predictor forecasts the future output based on present input and output of building structure. The controller is PID type whose parameters are yielded by neural networks self-tuning algorithm. Computer simulations show displacements of single and multi-story structure applied to NNPPID system about disturbance loads-wind forces and earthquakes.