The elevator group control system is a problem where new approachesare being used to optimize the cab assignment problem. Soft Computing methodscan be useful to assist the existing dispatching algorithm, predicting the passen-gers stop floor, or detecting the type of traffic pattern (up-peak, down-peak, inter-floor). In this work, neural networks has been used for predicting from where isgoing to come the next hall call and then this information is used to park the cabsadequately. An evaluation is carried out, using Dynamic Sectoring algorithm anddifferent service time analyzed. The results show that service level can beimproved using neural network for the demand prediction.
展开▼