With rapid economic development and an increasing number of high-rise buildings, the number of elevator units and service floors has also increased. How to use reasonable scheduling algorithms to reduce the energy consumption of elevator groups and improve operation efficiency has become a current research hotspot. To solve the problem of unsatisfactory performance of traditional group control algorithms in terms of time and energy consumption, an elevator group control algorithm is proposed, which implements an interactive artificial bee colony (ABC) for a new intelligent building platform. First, the Markov chain is used to simulate passengers’ call behavior. Second, three performance indicators, average riding time, average waiting time, and energy consumption, are selected to construct a multiobjective function. Finally, the traditional ABC algorithm is improved. After being deployed to an intelligent computing node, the interaction between the optimal solution and the optimal fitness value between multiple nodes can be realized so that each intelligent node can quickly and accurately solve the optimal dispatching plan. The experimental results show that under the condition of equivalent energy consumption, the interactive ABC algorithm reduces the average waiting time by 24.3% compared with that of the traditional ABC algorithm, improves the operation efficiency, and verifies its effectiveness and feasibility in elevator group control scheduling.