Minor component analysis (MCA) is a powerful tool in the signal processing field. Up to now, there are few algorithms, which can extract multiple minor components from input signals, and many limitation conditions exist before using some existing algorithms. In order to solve these problems, the M\"oller algorithm, which can only extract one minor component, is modified into a multiple minor components extraction algorithm by using the weighted matrix method. The proposed algorithm has no limitation on the smallest eigenvalue and has a good convergence property, while no norm operation is needed. Simulation results show that the proposed algorithm can efficiently extract the multiple minor components of an input signal and has a faster convergence speed than some existing algorithms.