Density-based methods have appeared as a practical approach for the clustering of data streams. While various density-based algorithms have recently been developed for data stream clustering, these algorithms are not without problems. The efficiency of the clustering is considerably reduced when an insufficient distance function is used. In addition, the majority of the approaches is their non-autonomous nature, which means they need manual tuning of their internal parameters. Unfortunately, the tuning process requires considerable effort and parameter results might become invalid after a certain time due to statistical changes in data or concept drift characteristics. In this study, we propose a new Density-based method for Clustering Data stream using Genetic Algorithm (DCDGA). This method using a Genetic Algorithm (GA) to adjust suitable parameters for the cluster radius and minimum density threshold to cover the density clusters more accurately. A Chebychev distance function is also introduced to calculate the distance between the Core Micro-Clusters (CMCs) center and the arriving data point. The proposed method was evaluated using an artificial and real dataset with different evaluation metrics. The experimental results were compared with another online density-based clustering. The recommended method provides an effective solution for improving the quality of clusters.