| Due to the advancement of communication network and necessity of real-time performance and distributed control systems in industrial control field, networked control has been soon prevalent in many control applications. Networked control systems (NCS) can be defined as feedback control systems including actuators, sensors and controllers, the operation of which is coordinated via a communication network. Employing the communication network in the control loop provides many benefits, such as the ease of maintenance and installation, the large flexibility and the low cost. Further more, a NCS supports various network topology architectures so that it is more integrated and robust.The defining feature of the NCS is that control information is exchanged using a network among control system components, which brings convenience to system design. It also raises several challenges, such as networked-induced delay and data packet loss, which can degrade system performance and cause unstability.This paper studies a networked system identification problem. Firstly, the two factors, namely, the networked induced delays and packet loss may cause the sampled data are non-synchronized non-uniformly(NSNU). In addition, the running mode of the actuator may impact identification method.Secondly, a generalized linear least square algorithm is proposed to solve the non-uniformly sampled data. This algorithm can identify continuous model parameters directly and produce unbiased estimation with very little computing time. Also, system initial parameter values can be identified.Lastly, the paper studies the identification of both system parameters and model order based on the NSNU data. The specific procedure is as follows:A subspaced method is introduced to estimate the model order. Next, use recursive simplified refined instrumental variable method to identify system parameters. Theory analysis and simulation examples are presented to demonstrate the effectiveness of the algorithm. |
PDF Full Text Request