Algorithm 1: Cluster Algorithms

Step 1: CH scores determine the number of categories

Step 2: FCM algorithm for cluster analysis

Input: Truck braking behavior data $X=\left\{x_1,x_2,\cdots ,x_N\right\}$, number of categories K, and threshold terminating iterations $\epsilon$.

Initialization: Take the random value of [0, 1] to initialize membership degree matrix U0; assume that the initial value of the number of iterations is h = 1.

Iterations: Solve the cluster center based on Equation (1).

Solve the new membership degree based on Equation (2).

Solve the objective function based on Equation (3).

h = h + 1.

Conditions for terminating iterations: $\left|J^{h+1}-J^h\right|\le \epsilon$, where $\epsilon$ is usually 0.0000001.

Output: Cluster results