Particle mixing is a very important process in many industrial reactors. It is important to consider how to quantify the particle mixing behavior accurately under different conditions. In this paper, the drawbacks of traditional mixing indexes, such as Lacey index are discussed, and then a novel particle mixing index, without grid dependency and particle color dependency was proposed based on the neighbor distance method and coordination number concept. Its characteristics such as parameters selection, calculation method and multi-direction were discussed in details. The proposed mixing index was used for investigating particle mixing behavior in the multiple-spouted bed under conditions of different gas velocity ratios and different particle density ratios. The experimental results have been used to validate the simulation results at first. It is found that the improved mixing index is more superior and effective compared with the traditional mixing index methods, and suitable for the investigation of particle mixing behavior. It was indicated that the density ratio is a key factor to influence particle mixing process. The novel particle mixing index without grid dependency and particle color dependency is recommended to study particle mixing behavior in other particulate processes in the future. (C) 2018 Elsevier B.V. All rights reserved.