Study On Control Strategy For Multi-Degree-of-Freedom Hydraulic Servo Systems

In a multi-degree-of-freedom (MDOF) hydraulic servo system (HSS), thereexists a strong coupling phenomenon among multi-channel hydraulic servo cylindersubsystems when they are acting simultaneously. The coupling directly influences thecontrol performance of the MDOF HSS. Deep study on the control method for theMDOF HSS is of great significance both in theory and application.This research aims to study an "Iron Bird" simulating an airplane in a lab. In theIron Bird, the rudder dummy system, which controls flying direction of the airplane,is a very typical and complex MDOF HSS. Meanwhile, the rudder dummy is stronglyinfluenced by strong nonlinear air currents, vary loads and other disturbances. It playsa tremendous role on finding a solution to the control for the MDOF HSS onstrengthening safety index as well as improving flying quality of the airplane.Some main research achievements are obtained as follows:Firstly, the developments of the control strategies for single-/multi-degree-of-freedom HSS were surveyed. Theoretical and applied researchdevelopments on the decoupling control for multivariable system, especiallydecoupling control strategies for the MDOF HSS, were summarized in detail.Secondly, a fuzzy sliding mode variable structure control was researched andapplied to a HSS. Although sliding mode variable structure control (SMVSC) hasexcellent robustness against external disturbance and parameter variances, there existschattering. A solution to chattering reduction is also a problem related to hydraulicservo control. A quasi-sliding mode variable structure control (quasi-SMVSC) basedon fuzzy logic theory and boundary scheduling technique of quasi-SMVSC wasdeveloped. It can fuzzily and dynamically adjust the boundary width of sliding-mode.Consequently, strong robustness against disturbances due to high-frequency controlswitching, static state performance and response speed of the HSS are guaranteed.Chattering from high-frequency switching control is controlled within the permissiblerange so that the HSS can work normally.Thirdly, to balance the design conflict of SMVSC between chattering reductionand high-precision control, a SMVSC based on coordination optimization algorithmwas presented. Steady-state error and control switching frequency are used as thesystem performance indexes in the optimization. By using the boundary layer widthtuning law based on steady-state error and adopting a coordination optimizationgenetic algorithm, a boundary layer width tuning rate is optimized and added to thenonlinear control term of SMVSC. An optimized SMVSC for HSS was realized. Theproposed control successfully solved the problem of design conflict. That is, thedesign conflict between high steady-state precision and softened control action in thehydraulic servo sliding mode variable structure control system are well balanced, andgood comprehensive performance can be obtained. Without decreasing the robustnessof the system, the high-frequency switching control can be softened to the mostextend and the expected control accuracy is to be realized. If combining sliding modewith co-ordination optimization technique, the tuning of boundary layer width is more effective to optimally coordinate chattering reduction as well as high-precision controlin SMVSC. Especially, this method plays a predominant role on adapting controlactions to the practical control constraints of the HSS.Fourthly, two simulation techniques which are hardware-in-the-loop (HLP) andcollaborative simulation respectively were studied. Based on HOPSAN andMATLAB/SIMULINK, a method of collaborative control simulation was presented.The simulation results verify the effectiveness of the collaborative control simulation.It can interconnect hydraulic, mechatronics, control theory and control engineering,and merge them together into a hybrid heterogeneous hierarchical control simulationsystem. This system can collaboratively and distributively operate according to somerules, and produce modeling and control results. Eventually, this method results inmulti-field simulation technology called collaborative control simulation technique.Meanwhile, a platform is set up for effectively, economically and reliably dealing withnonlinear modeling and optimal control issues of HSS. It is of great significance toreduce the developing cost for the control system of large-scale complex HSS and toimprove the working efficiency.Fifthly, this dissertation demonstrated the theoretical basis of the decouplingcontrol for the MDOF HSS. Based on the coupling relationship among differentchannels of the MDOF HSS, a basic idea of comprehensive decoupling control tosolve the coupling in the MDOF HSS was proposed.Sixthly, a physical model, a HOPSAN model as well as a MATLAB/SIMULINKmodel were set up for the MDOF HSS of the "Iron Bird" in this paper. This sets asolid basis for further study on the control for the MDOF HSS.Seventhly, a comprehensive control was proposed for the rudder dummy of theMDOF HSS of the "Iron Bird". Before designing the multi-channel decouplingcontroller, interactions among multi-channel were neglected. Main stable SMVSCwas designed for each channel to stabilize the MDOF HSS of the "Iron Bird". Then,based on Baumgarte's decoupling control design method and approximate systemmodel, a feedforward plus state feedback decoupling controller was designed andapplied to the decoupling control for the MDOF HSS of the "Iron Bird". Ultimately,by considering the influence of un-complete decoupling on the MDOF HSS of the"Iron Bird" as the external disturbances d (d₂₁, d₁₂) to the two single channels, acollaboration optimization algorithm was constituted to improve the control precisionas well as to strengthen the robustness against disturbances of the single-channelSMVSC system in the MDOF HSS. By using the proposed SMVSC based oncollaboration optimization algorithm, the influence of un-complete decoupling on theMDOF HSS of the "Iron Bird" was overcomed.In the last part of the dissertation, the whole research conclusions and futureresearch directions were presented.