Research On Multi-objective Dispatching Planning Of Elevator Group

As the great development of electronics, computer application and network communication, the technology of Elevator's Real-time monitoring has been developed greatly in recent years. Therefore, the EGCS (elevator group control system) can get each elevator's dynamic state more overall and accurately, which creates more favorable conditions for the artificial intelligence technique, and also provides more effective ways for improving the EGCS's performance.The elevator group control dispatching method is the core of the EGCS, and it directly influences the operating of each elevator and the service quality of the elevator system. It's a decision problem about a complex stochastic system which is nonlinear, uncertain and multi-objective, and the Genetic Algorithms (GA) is a new solution to the complex problem which is uncertain, nonlinear and irreversible, so this thesis tries to use the GA to solve this decision problem.For intensive study, it's needed to make a model of the EGCS. By analyzing the EGCS, the main controlled objectives are presented, including the waiting time, riding time and crowding ratio of passengers and the EGCS's power consumption, and a multi-objective planning EGCS model is built.After studying the GA in multi-objective planning, a logical approach using GA is found to the elevator group control dispatching problem. According to the actual conditions, the design of each logical approach is introduced detailedly, which contains coding method, population initializing, fitness function design, genetic operators design, parameters selecting and end marking. And a multi-objective GA process is got to solve the elevator group control dispatching problem. Then simulation software is written for analyzing, and the feasibility of solving the elevator group control dispatching problem by using the GA is proved. Finally, it's simulated in virtual environment, and the results show that the method is effective.