Study On Tracking Performance Limitation For Feedback Control Systems Over Networked Environment

The performance research for control system has attracted more and more attentions of scholars, there have been a series of important results for the performance analysis of the traditional feedback control system, and these findings may provide an important theoretical basis for the analysis and design of actual systems.With the rapid development of technology and the accelerating modernization, information and network are more and more popular, and it will be difficult for the traditional control system to meet the needs of the information society, the network control system appears everywhere in our real life. Because of the involvement of networks, there have been enormous changes in our lives, the network brings much convenience to our lives but also has brought great challenges for the control system analysis and design to some extent, and the system performance significantly deteriorated. Currently the analysis and discussion for networked control systems mainly focus on the stability analysis of the system, while the network performance factors have not been considered, but the performance analysis of the network control system, especially the study of performance limits, is particularly important. This article is to study the optimal tracking performance for networked control systems based on the communication parameters constraints, aiming to reveal the intrinsically relationship between the tracking performance and the internal structure of the system and the characteristics of the communication parameters. The main contents are summarized as follows:The optimal tracking performance problem of single-input single-output feedback control system with limited energy constraints has been studied. There has been Gaussian noise in the feedback channel and random disturbance signal in the forward channel, while the channel transmission of energy is limited. The optimal tracking performance is obtained, based on frequency domain Euler parametric approach to design a suitable controller. The results reflect that the additive white Gaussian noise and intrinsic characteristics of the plant, as well as disturbance input signals may degenerate the tracking performance. The intrinsic link between the tracking performance limit with the charged object structure and network noise is explored.The optimal tracking performance of the linear time-invariant continuous system in the presence of model uncertainty, at the same time the feedback channel with quantization control input infect, is obtained. Not only the model uncertainty is considered, but also the network channel is suffered the quantized control input. The two-degree-of-freedom controller structure is adopted. The study results reveal that the optimal tracking performance of the control system is depend on the quantization noise of the communication channel and the uncertainty error.The optimal tracking performance for networked control systems, in the presence of Gaussian noise of the forward channel and feedback channel with quantization control input infect, is studied. And the codec is introduced the feedback channel to reduce the quantization error. The optimal tracking performance index is obtained and the results show that the optimal tracking performance of the control system depends on the network noise and the communication noise error.The optimal tracking performance of the linear feedback is studied, in the presence of model uncertainty, at the same time there exist communication noise in both feedback and forward channel. The model uncertainty and the network channel noise are both considered, two-degree freedom controller structure is adopted, and the best tracking performance is obtained. The results show that the network channel noise and intrinsic characteristics of the plant, as well as model uncertainty part may degenerate the tracking performance.Averaged tracking performance limitations in linear time invariant multi-input multi-output systems with random signal reference input has been studied. The random input signal considered is the Brownian motion sequence. The case of the uplink channel and downlink channel bandwidth constraints and channel noise, by means of two degrees of freedom controller structure, and all-pass decomposition and decomposition inside and outside, the optimal tracking performance of the network control system for random signals. The results show that the optimal tracking performance of the system depends not only on the inherent characteristics of the controlled object, such as non-minimum phase zeros, unstable poles, and quantitatively reveal how the channel bandwidth and noise impact on the tracking performance of the network control system.Finally, a summary for all discussions is given in the dissertation. Future works that related to this work are also presented.