Artificial bee colony(ABC) algorithm performs well in the exploration process, but develops poorly in the exploitation process. How to balance the exploration and development of the algorithm during the evolution process has always been a challenge for the ABC algorithm.At the same time, for complex problems, the ABC algorithm may experience premature convergence or stagnation during the evolution process. To overcome these challenges, this paper proposes an effective variant of the ABC algorithm (random Neighborhood search Artificial Bee Colony with Population State Evaluation), referred to as NABC-PSE. In the NABC-PSE algorithm, a random neighborhood method is introduced, and different search strategies are designed for the employed bees and onlooker bees phases based on the best solution within the neighborhood. Additionally, a population state evaluation mechanism is added during the onlooker bee phase to allocate appropriate evolutionary strategies for the population when it becomes stagnant or converges prematurely. To address the issue of the roulette wheel mechanism where individual fitness becomes less distinct in the later stages of evolution, a selection strategy based on sorting is employed. Finally, to make more effective use of dimensional information, a greedy selection method is used to choose the dimensions for improvement.To validate the effectiveness of the algorithm, the NABC-PSE algorithm was tested on a set of 22 benchmark functions and compared with some other ABCs and several state-of-the-art algorithms. The results show that the proposed algorithm has high solution quality, fast global convergence rate.