The paper introduces an advanced pattern recognition strategy for classifying various robotics during curve negotiation tasks. The method combines k-means clustering and Support Vector Machine (SVM) techniques to enhance the classification process. Initially, k-means clustering is used to segment robot data into distinct groups, reducing the number of support vectors and improving feature discrimination. Subsequently, the SVM method constructs a discriminative hyperplane to classify the robots accurately. The effectiveness of the k-SVM approach is validated through cross-validation experiments, demonstrating superior performance compared to traditional SVM methods in robot group classification. The study aims to improve collaboration efficiency, minimize emissions, and integrate specific robot factors into control strategies, making it suitable for multi-robot systems. The proposed method is particularly effective in handling non-linearly separable datasets, as evidenced by its lower classification errors and faster training times.The paper introduces an advanced pattern recognition strategy for classifying various robotics during curve negotiation tasks. The method combines k-means clustering and Support Vector Machine (SVM) techniques to enhance the classification process. Initially, k-means clustering is used to segment robot data into distinct groups, reducing the number of support vectors and improving feature discrimination. Subsequently, the SVM method constructs a discriminative hyperplane to classify the robots accurately. The effectiveness of the k-SVM approach is validated through cross-validation experiments, demonstrating superior performance compared to traditional SVM methods in robot group classification. The study aims to improve collaboration efficiency, minimize emissions, and integrate specific robot factors into control strategies, making it suitable for multi-robot systems. The proposed method is particularly effective in handling non-linearly separable datasets, as evidenced by its lower classification errors and faster training times.