Energy-Saving Elevator Group Control Intelligent Method Research Based On Multi-agents

With the requirement increasing of energy-saving, it becomes the focus of society. The study of the multi-Agent system has become the new hotspot research direction with the closely combining of the artificial intelligence and energy-savingBased on the basic theoretical study of multi-Agent system, this paper will focus on the application of the group dispatching which make the system more flexible. It designed the multi-Agent group model base on the multi-Agent model. And it determined the multi-objective evaluation function of the group control multi-Agent, each of the objective function of linear weighted combined the energy-saving features into the evaluation of adaptive function.It designed the dispatching method of the multi-Agent group control by applying the multi-Agent coordination and negotiation techniques. Elevator dispatching problem will be understood the process of coordination and consultation through the several elevator unit to handle the random calling task. This approach was different from the group control dispatching tasks assigned by a group controller to complete, it made an Agent for each elevator unit, each agent was calculated according to the objective function, reasoning, decision-making, through mutual coordination between the lift to complete the best of the dispatching.By the comparison between the evaluation index of the multi-Agent group dispatching with the evaluation index of the general dispatching, it shows that energy saving advantage of the multi-Agent group control techniques.