Quantized Feedback Stabilization Of Networked Switched Systems With Finite Bit Rates | Posted on:2023-09-03 | Degree:Doctor | Type:Dissertation | Country:China | Candidate:R Chen | Full Text:PDF | GTID:1528306902453714 | Subject:Control Science and Engineering | Abstract/Summary: | PDF Full Text Request | Recently,switched systems play an important role in the field of hybrid systems and have become a hot topic in the control field.Switched systems have been widely applied to control systems whose dynamics may exhibit instant variations.In reality,switched systems may transmit their feedback signals through a digital communication network.Since the network can only provide a finite network bandwidth,feedback information can only be transmitted at some discrete instants and needs to be quantized into a finite number of bits,which causes that most of the conventional results on switched systems cannot be applied directly.Therefore,research on quantized feedback stabilization of networked switched systems with finite bit rates has caught more and more attention.The main concern of this research is to design some encoding and control strategies to stabilize switched systems at a finite bit rate.For the purpose of saving network resources,the required stabilizing bit rate for the concerned switched systems is expected to be as low as possible.Compared with periodic sampling,eventtriggered sampling can save network resources efficiently.Therefore,the design of event-triggered sampling strategies is also a main concern of this thesis.Moreover,compared with traditional control systems,the dynamics of switched systems may exhibit instant variations,which makes the design of event-triggered sampling strategies more challenging.The main contents and contributions of this thesis can be summarized as follows:1)It proposes encoding and periodic sampling control strategies to achieve the asymptotic stability of a switched system at a finite bit rate,and analyzes the effects of network delay on the system’s stability.It considers the coupling effects of network delay,mode switch and quantization errors on the system’s stability,and derives a lower bound on the average dwell time and sufficient bit rate conditions to ensure the desired stability of the concerned switched system.Moreover,it quantitatively analyzes the effects of network delay on the sufficient stabilizing bit rate conditions.2)For saving network resources,it proposes a mode-dependent continuous-time event-triggered sampling strategy.An estimation model is established at both the encoder side and the controller side to generate a real-time estimate of the system state.Based on the state estimation error,it proposes a continuous-time event-triggered sampling strategy to reduce the feedback bit rate.Under the event-triggered sampling strategy,only if the state estimation error reaches a pre-defined threshold,the system sampies and transmits feedback signals.Moreover,by extracting the extra information from the event-triggering time instants,it proposes a novel encoding strategy based on a traditional encoding strategy.Compared with the traditional encoding strategy,the novel encoding strategy consumes less feedback bits.Under the event-triggered sampling strategy,it gives sufficient stabilizing bit rate conditions for the concerned switched system and proves that the required stabilizing bit rate is strictly lower than that of existing literature.3)Based on the work above,it studies the effects of model uncertainty,network delay and external disturbances on the system’s stability and proposes a mode-dependent periodic event-triggered sampling strategy.By applying reachable set approximation and propagation approaches and increasing the upper bound of the state estimation error,it can well handle with the effects of network delay and model uncertainty on the system’s stability and the problem of mode mismatch between the encoder side and the controller side caused by the unknown mode switch time.It also applies the proposed encoding strategy and quantitatively analyzes the effects of network delay and model uncertainty on feedback bits saved by the novel encoding strategy.Moreover,in order to save network and computing resources,it first designs a discrete time sequence whose intervals are no shorter than the upper bound of network delay,and then based on this time sequence proposes a periodic event-triggered sampling strategy.Under the event-triggered sampling strategy,the system verifies the event-triggering conditions only at the discrete time instants in that time sequence to ensure that even in the existence of network delay,the controller can still be aware of the event-triggering time instants.Furthermore,it gives sufficient bit rate conditions to ensure the input-to-state stability of the switched system and quantitatively analyzes the effects of network delay and model uncertainty on the required stabilizing bit rate for the switched system.4)It studies the effects of impulses on the system’s stability.The frequent occurrence of impulses may hurt the system’s stability and even destabilize the system.Similar to the approach of restricting the occurrence frequency of mode switches,it introduces dwell time and average dwell time of the impulsive signal to restrict the occurrence frequency of impulses.Moreover,it considers the coupling effects of impulses,mode switch,network delay and external disturbances and derives the upper bounds of the state estimation error for both the case with neither mode switch nor impulse in a sampling interval and the case with one switch or one impulse in a sampling interval.Based the upper bounds of the state estimation error for above two cases,it proposes a periodic event-triggered sampling strategy to save the occupied network bandwidth.Furthermore,it gives sufficient bit rate conditions to ensure the input-to-state stability of the impulsive switched system and proves that when network delay is small enough and the switching signal and the impulsive signal satisfy a coupling average dwell time condition,the proposed control strategies can ensure the input-to-state stability of the impulsive switched system at a lower bit rate than existing literature. | Keywords/Search Tags: | Switched systems, Networked control systems, Finite bit rate, Event-triggering, Network delay, Model uncertainty, Impulsive switched systems | PDF Full Text Request | Related items |
| |
|