Compensation Of The Delay & Packet Dropouts In CAN Network And Its Application In Hybrid Excavators

Controller Area Network (CAN) is a typical fieldbus wildly used in cars, construction machineries, industrial robots and other industrial applications. The communication in CAN has random delays. When the CAN network is suffering from communication errors due to electromagnetic interferences, cable aging and loose connections, the communication delay will increase and lead to packet dropouts. Since most of the key components of the equipments are controlled through CAN network, the communication delay and packet dropouts will degrade the control performance, and even result in equipment damage. If the signals affected by the delay and packet dropouts can be compensated, the control performance can be improved. Thus, the equipments can be operated reliably, safely and efficiently under communication errors. Meanwhile, the down time of equipments can be reduced and the equipment maintenance time can be flexibly determined.CAN network is sensitive to bus load due to its communication mechanism. The communication quality will degrade as the bus load increases. Therefore, the compensation of delay and packet dropouts should avoid to increase the bus load of CAN network. In addition, as the communication errors randomly occurred, the delay and the number of consecutive packet dropouts vary widely, and have unpredictable probability distributions. In order to improve the control performance as far as possible, the compensation method should be able to compensate the delay and packet dropouts according to their exact value. However, the current compensation methods cannot satisfy the aforementioned requirements at the same time. Therefore, it is necessary to study on novel compensation methods for the communication delay and packet dropouts so as to guarantee the CAN-based control system to perform well under communication errors.In this dissertation, we propose a Distributed Compensation Scheme when taking the characteristics of CAN network into consideration. Based on the monitoring and analysis of the communication activities on CAN bus, a real-time delay estimation method, which won’t affect the bus load of CAN, is proposed. By using the Electric Swing System (ESS) of a hybrid excavator as the research object, the Distributed Compensation Scheme for Output Feedback Control System and State Feedback Control System is studied respectively. In the light of limited compensation ability of the Distributed Compensation Scheme under serious communication errors, an online health monitoring method for CAN network is proposed. The methods proposed in this dissertation can work together to guarantee the control system to be operated safely and efficiently while not affecting the communication quality of CAN. As a typical Networked Control System (NCS), the proposed methods can also provide references for the NCSs based on other networks.The dissertation is organized as follows:In chapter 1, the significance of study on the compensation of communication delay and packet dropouts under the background of control system networking is discussed. In the introduction of the state-of-art of the delay compensation in NCSs, the characteristics and existing problems of several popular schemes are analyzed in detail. In the introduction of the state-of-art of the delay detection in NCSs, the characteristics and existing problems of current methods are analyzed. Finally, the main research contents of the dissertation are presented.In chapter 2, the compensation scheme for the communication delay and packet dropouts in CAN-based NCSs is studied. Firstly, we analyze the characteristics of the communication delay and packet dropouts and how they occur in CAN network. Then, the essential impacts of the delay and packet dropouts on the control performance are studied. Finally, based on the aforementioned researches, the Distributed Compensation Scheme for communication delay and packet dropouts is proposed in the light of limitations of CAN network, and the common issues of compensator design is discussed.In chapter 3, due to the difficulties in the calculation, prediction and measurement of the delay and packet dropouts in CAN, a new on-line extraction method for the communication delay and packet dropouts in CAN is studied. Firstly, based on the analysis of communication activities on CAN bus, the main idea of delay estimation and packet dropouts detection is discussed. Secondly, in the light of the difficulties in identification of the uncertain delay components, the message traces on CAN bus is analyzed in depth, and then a novel method for the online estimation of communication delay and the online detection of packet dropouts is proposed. Finally, an experimental platform, which is used to imitate the practical CAN system of the hybrid excavator, is constructed, and the effectiveness and general applicability of the proposed method are studied.In chapter 4, based on a linear SISO system, the realization of the proposed Distributed Compensation Scheme in Output Feedback Control System (OFCS) and in State Feedback Control System (SFCS) is studied in light of the characteristics of different control structures. Firstly, the design of the feedback compesator based on the plant model and the forward compensator based on the control law are developed for the OFCS, considering that the OFCS calculates the control command based on the feedback system output which can be obtained in real time by the actuator. Then the design of the feedback compesator based on the state obsever and the forward compensator based on state-sending or independent-state-estimation are-proposed for the SFCS considering that the state obsever is in the controller. Finally, the simulation model is established in MATLAB/simulink and the compensation effect of the proposed method is researched.In chapter 5, the proposed Distributed Compensation Scheme is applied to the electric swing system (ESS) of hybrid excavators to improve the operability, comfort, reliability and safety of the hybrid excavator under communication errors. Firstly, in light of the time-varying rotary inertia of the swing platform and the difficulty to identify the detailed parameters of the ESS, a siplified mathematical model of the ESS, which can reflects the main dynamics of the ESS, is established, and the corresponding model parameters are identified based on the recursive least square algorithm with forgetting factor. Then the realization of the proposed Distributed Compensation Scheme in the ESS, including the feedback compensator and the forward compensator, are presented. Moreover, in light of the limited compensation ability of the Distributed Compensation Scheme under serious communication errors, an online health monitoring and alerting method for the ESS is proposed and validated experimentally. Finally, the approch is emmbedded in the practical ESS and validated experimentally.In Chapter 6, the main conclusions and achievements are summarized and the further research work is put forward.