Abstract: Compared with the classical hidden Markov model, the hidden quantum Markov model has the advantages of fast solving speed and fewer parameters, which has attracted much attention. But in the process of transition from classical to quantum, we find that the hidden quantum Markov process is closely related to the quantum open system. Different from previous studies, this article starts from the open quantum system, studies the relationship between the hidden quantum Markov and the main equation corresponding to the open system. On this base, taking the quantum transport system as an example, the relationship between the quantum conditional master equation and the hidden quantum Markov model is theoretically revealed. And then a learning algorithm based on the idea of maximum likelihood estimation is proposed to solve the parameter solving the problem in the hidden quantum Markov model.