Studies On The Optimization Of The EGCS

With the prevalence of modern intelligent buildings,People's demands to the elevator system functions are also more and more high.A single elevator cannot satisfy people's demands for transportation capability.The thesis improves the intelligent scheduling algorithm of the EGCS based on Adaptive Multi-objective Scheduling Algorithm regarding the Elevator Group Control System as the research object,decreasing congestion degree of lift car as the goal after the EGCS,the Fuzzy Neural Network as well as the Reinforcement Learning is studied.,after that,The thesis also simulates the EGCS.First,the structure of the EGCS is given;the characters and the targets of performance evaluation for the EGCS are studied.By taking the shortest average waiting time of passengers,the minimum consumption and the least crowded degree as the evaluate indexes.The comprehensive evaluation function is established with the weighted linear combination method.Then the architecture of self-adaptive multi-objective optimization is built,the Reinforcement Learning method is applied to adjust the evaluation function parameters;Adaptive Multi-objective Scheduling Algorithm of the EGCS is completed.Finally,using the Matlab/Simulink simulation software,the Traffic Flow Simulation Model,the Elevator Running Simulation Model,the Single Elevator Control Simulation Model and the Elevator Group Control Simulation Model are founded.Taking a three sets 12-story elevator control system as the example,the Adaptive Multi-objective Scheduling Algorithm and the Multi-objective Programming Algorithm are applied to simulate the EGCS separately.The simulation results which can reflect the service quality is obtained.The researches in this paper shows that the Adaptive Multi-objective Scheduling Algorithm applied in the EGCS can enhance the adaptation of the system effectively as well as improve the performance of the EGCS.It can also provide a reference to the similar optimization problem.