In This Paper We Join the Biggest Least Distance Algorithm and the Conventional K-Means Algorithm to Propose a Further Developed K-Means Clustering Algorithm. This Further Developed Algorithm Can Make Up the Weaknesses For the Conventional K-Means Algorithm to Decide the Underlying Point of Convergence. the Further Developed K-Means Algorithm Adequately Tackled Two Detriments of the Conventional Algorithm, the First Is More Noteworthy Reliance to Decision the Underlying Point of Convergence, and Another Is Not Difficult to Be Caught In Neighborhood Minimum[1][2].