http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Introducing Prior Knowledge for a Hybrid Accident Prediction Model
Ali Mansour Khaki,Abolfazl Karimpour,Hadi Sadoghi Yazdi 대한토목학회 2017 KSCE JOURNAL OF CIVIL ENGINEERING Vol.21 No.5
This study aims to address two potential issues regarding accident prediction models. (a) What are the benefits of using the prior information on the accuracy of prediction model, and (b) how to include the available prior knowledge of accident occurrences data in the prediction process? In accident databases, the prior knowledge can be defined as the most probable points in which an accident has happened in the earlier years and is closely correlated to its upcoming years. Large databases such as traffic and accident databases, inevitably contain noisy data. Therefore, to have accurate results, using approaches to alleviate the impact of these anomalies is significant. In this research, a hybrid method based on a module of prediction and a prior knowledge block is proposed. The module used for prediction is Recursive Least Square filter and Maximum a Posterior (MAP) estimator is used as the prior knowledge block. Results indicate an increase in accuracy of prediction by using the proposed hybrid model.
Sholeh Yasini,Mohammad Bagher Naghibi Sistani,Ali Karimpour 제어·로봇·시스템학회 2015 International Journal of Control, Automation, and Vol.13 No.1
This paper develops a concurrent learning-based approximate dynamic programming (ADP) algorithm for solving the two-player zero-sum (ZS) game arising in H∞ control of continuous-time (CT) systems with unknown nonlinear dynamics. First, the H∞ control is formulated as a ZS game and then, an online algorithm is developed that learns the solution to the Hamilton-Jacobi-Isaacs (HJI) equation without using any knowledge on the system dynamics. This is achieved by using a neural network (NN) identifier to approximate the uncertain system dynamics. The algorithm is implemented on actor-critic-disturbance NN structure along with the NN identifier to approximate the optimal value function and the corresponding Nash solution of the game. All NNs are tuned at the same time. By using the idea of concurrent learning the need to check for the persistency of excitation condition is re-laxed to simplified condition. The stability of the overall system is guaranteed and the convergence to the Nash solution of the game is shown. Simulation results show the effectiveness of the algorithm.