The dynamics of hydraulic systems are highly nonlinear; examples like the strongly coupled multiple-input-multiple-output (MIMO) mechanical linkage dynamics, the deadband of control valves, and the nonlinear flow mapping. Aside from the nonlinear nature of hydraulic dynamics, hydraulic systems also have large extent of model uncertainties, either due to parametric uncertainties (e.g., large variations of load and hydraulic parameters such as bulk modulus) or uncertain nonlinearities (e.g., leakage flows and frictions). These nasty characteristics make the precision control of hydraulic systems rather difficult, and there has been a significant lack of advanced controls that deal with all these issues well and systematically.
As a stepping stone toward the systematic design of high performance control algorithms for hydraulic systems, a nonlinear model based adaptive robust control approach is presented and carefully examined through both rigorous theoretical analysis and experimental verifications. Specifically, nonlinear physical model based analysis and design is used to address the inherent nonlinearities and strong coupling of hydraulic dynamics. Adaptive robust control (ARC) is applied to deal with various model uncertainties effectively; the approach uses fast robust feedback to attenuate the effect of various model uncertainties as much as possible for a guaranteed performance in general and uses learning mechanism such as parameter adaptation to reduce model uncertainties for a better performance. Backstepping design via ARC Lyapunov functions is adopted to address the issue of unmatched model uncertainties.
The dissertation starts from the precision motion control of one DOF hydraulic servosystem driven by either a double-rod/double-actuating or a single-rod/double-actuating hydraulic cylinder with constant or time-varying unknown inertia. The stability proof of the zero tracking error dynamics associated with single-rod hydraulic cylinders is given. Swing motion control of a three DOF robot arm driven by single-rod hydraulic actuators with the other two joints fixed or actuated simultaneously are used as experimental case studies.
For systems that use cheaper proportional directional valves instead of servo valves, addressing strategies are presented to deal with the nonlinear valve characteristics such as deadband and nonlinear flow gain coefficients. Different methods are presented to improve the response of sluggish proportional directional control valves.
The coordinated motion control of multi DOF electro-hydraulic robotic arm is then studied. Two methods are proposed to avoid the need of acceleration feedback for ARC backstepping designs; one uses a nonlinear ARC observer to recover the state needed for the ARC backstepping design, and the other makes full use of the physical property that the adjoint matrix and the determinant of the inertial matrix could be linearly parametrized by certain suitably selected parameters and employs overparametrizing method.
An integrated direct/indirect ARC algorithm is also constructed to meet the dual objectives of good tracking performance and parameter estimations. The resulting accurate parameter estimates may be used for other purposes such as machine and component health monitoring and fault detection.
This work addresses the design and experimental implementation of controllers for hydraulic actuators trajectory tracking. The hydraulic actuator mathematical model is interpreted as a mechanical subsystem driven by a hydraulic one. Using this interpretation, cascade controllers are presented based on linear and nonlinear models that do not include the valve dynamics. Analyzing the hydraulic subsystem stability, one demonstrates the limitation caused by the valve dynamics, and a cascade controller based on a model that include the valve dynamics as a first order system is proposed. The proposed controller is synthesized by using Lyapunov direct method and the closed loop is exponentially stable when the actuator parameters are known. The cascade controllers are experimentally implemented on a hydraulic actuator composed of a proportional valve and a double-rod cylinder. In this implementation, the effects of dead-zone, friction and proposed compensations are discussed. Experimental results illustrate the main features of this work's cascade controllers.
Using Computational Fluid Dynamics (CFD), the thesis investigates the conditions under which cavitation forms within a 9 piston variable displacement axial piston pump. Four 3D CFD models of the pump and pressure compensator were successfully constructed using the porous mesh technique developed by the author. Split 3D models and a combined 3D model of the pressure compensation circuit were constructed and successfully validated against transient data. These models showed that cavitation forms at the metering edge of the spool under all operating conditions. Two further models of the full pump were constructed.
The first was run as a steady state speed model and indicated the formation of cavitation at both the inlet and outlet port leading edges.
The second model linked the pump model to the pressure compensation circuit data and provided a transient model of the pump operation. This model indicated that cavitation formed at the inlet and outlet port leading edges under all conditions. Cavitation was most evident when the piston stroke was at its maximum.
The method referring to the ISO standard concerned with the purpose of measuring effective area is frequently used to determine the flow rate characteristics of a test valve. This method is to measure the flow rate through a valve and differential pressure across the test valve. However, the experiment and calculation is complex and the accuracy of the method depends to a great extent on the precision of the flow meter. On the other hand, a simple method to measure the effective area is defined by JIS B 8373. In this method, compressed air is discharged from chamber through the test valve without using flow meter. The effective area is measured indirectly from the pressure response of air in the chamber. Due to discharging, the chamber air temperature drops about 40 [K], and the cool air enters the pipe but recovers because of the heat transfer between air and the pipe. This study shows the effect of connecting pipe air temperature recovery on the measurement of effective area. In this research, experiments and simulation were carried out to estimate the air temperature recovery effect. Generally, the connecting pipe effect caused by air temperature recovery is about 2% to 3%.
The constant connection between crankshaft and camshaft of conventional combustion engines is always leading to compromises in the design layout, which is making a negative impact on the system effectiveness. This is particularly true for the gas exchange of spark ignition engines, where the rate control is made by the throttle valve. Variable valve trains offer the possibility to control the incoming volume directly by the poppet valve.
This survey was focussing on the investigation of the driving power of an electrohydraulic linear drive, which is conceived to be used as a variable valve train for combustion engines, and in addition, to find ways how to optimize the amount of driving power.
Therefore first the different parts of the driving power where analyzed in theory. As a result of this, three different basic causes could be identified: During the movement incurring friction of different manner, conversion losses from potential to kinetic energy and hydraulic leakage flows in particular phases of the movement.
Each of this three areas under investigation was first of all theoretically analyzed by mathematical models, to permit the later interpretation of experimental results. Afterwards the existence of the different parts of the driving power was proved and quantified in specific experiments. Thereby also the validation of the theoretical model was performed.
Finally the results of the theoretical and experimental investigations were combined in a complete approach for an energetic optimized design of this specific drive. With this approach a reduction of the afforded driving power of more than 65% could be proved.
As result of this survey the different parts of the afforded driving power are now completely known. The experimentally proved theoretical models in combination with the developed approach are making possible a well directed optimization of the afforded driving power in the layout phase.
In this thesis real characteristics of a high-pressure tap-water cylinder drive were studied from the control-engineering point of view. Basic dynamic parameters and feasibility of force control in position control applications were studied experimentally and with a nonlinear model.
The influence of the Coulomb friction force was as much as half of the actual damping factor. The rest of the damping could be explained by viscous friction force and valve leakage. In practice, the damping factor has to be evaluated without any actual connection with the numerical values of physical parameters included in linear models. Pressure and velocity oscillations could be prevented efficiently using different profile generators. The influence of servo valve nonlinearities proved to be considerable but the use of nonlinear control-notch and load-pressure compensation could compensate for it. Force-control studies showed that direction control servo valves can be used in some force-control applications, but there are problems with varying volume and piston movement.
According to this study, similar design methods as in oil hydraulics can be used to design water hydraulic position servos. The main difference lies in the different characteristics of components, not in water as a pressure medium.
In this thesis the temperature behaviour of the piston cylinder assembly in swash plate type axial piston pumps is investigated. For the theoretical investigations a simulation model is developed to predict the temperature behaviour in the gap between piston and cylinder as well as in the rotating group. This model is based on a description of the non-isothermal gap flow between piston and cylinder. For this purpose the Reynolds equation, the energy equation and the equations of motion are simultaneously solved numerically. It is known that the piston in swash plate axial piston machines undergoes an eccentric micro-motion, which determines the final gap height for each time step. To calculate this motion a simplified equation of motion, based on external forces and the hydrodynamic forces generated by gap flow, is used. The temperature distribution in the rotating group is obtained with the help of a three-dimensional heat transfer model considering the energy dissipation in the gap, the heat convection and the heat conduction. The theoretical investigations were combined with extensive experimental investigations using a modified standard axial piston pump. Totally 60 thermocouples have been implemented in the rotating cylinder block. The temperature distributions in the whole cylinder block of the machine and the instantaneous pressure in the displacement chamber were measured under real operating conditions using a telemetry unit. The results were used to verify the simulation model.
In this dissertation, motion synchronization of multi-cylinder electro-hydraulic (EH) lift systems is investigated. Among the different methods of synchronizing the displacement of multiple cylinders in a hydraulic lift system with unknown but bounded load, EH control presents a set of attractive trade-offs. A general model for an n cylinder EH lift is developed and includes the vertical displacement, pitch and row motion of the load. It can be seen that without external synchronization control, the lift is an unstable system where the difference in cylinder displacements will increase under unbalanced load and results in significant pitch and row motion that will topple the load. Two prototype EH lift systems, a two post and a four post lift, are constructed to evaluate different motion synchronization control strategies. Linear Quadratic Regulator (LQR) and Cross-Coupled Controller (CCC) approaches are applied using the linearized lift model. Experimental result showed noticeable improvement over existing mechanical constraint solution. Due to the inherent nonlinear dynamics and uncertainties associated with the hydraulic system, a nonlinear perturbation observer is proposed to estimate the perturbation that combines the effect of parametric uncertainties and external disturbances. A nonlinear force control algorithm is then design to improve force tracking using the result of the perturbation observer. By exploiting the unique structure of the lift system, an inner and outer loop control is realized through a two-step back-stepping synthesis approach. The outer loop motion synchronization control is achieved using linear MIMO robust control techniques and the inner force control loop is achieved using the previously developed perturbation observer based nonlinear force control. The overall system stability is shown. Both simulation and experimental results verified the feasibility of the proposed two-loop control strategy. Experimental results showed that the proposed controller achieves steady-state synchronization down to the sensor resolution level as well as demonstrated improved transient performance compared to the linear designs. Valve dynamics are ignored in the current design. Sensor noise and valve dynamics are the two most significant factors that limit the performance of the system.
This thesis deals with a sliding mode control algorithm when it is applied to some representative hydraulic systems such as an electro-hydraulic servo system, a 3-D.O.F. hydraulic parallel mechanism, and an excavator. It is well known that, in such systems, a steady-state error can appear in time responses that originate from disturbances external to the systems. As a method for overcoming this problem, a disturbance observer is used in the sliding mode control algorithm. In the proposed algorithm, an input for the suppression of disturbance is estimated by the disturbance observer. It was verified by experiments and simulations that, in spite of the existence of unknown disturbances, the steady-state error and chatter phenomena were notably reduced by such use of a disturbance observer.
In manufacturing technology, it is a dominant tendency in recent years that machine tools for machining, grinding and so on, employ electric operated actuators such as a servo-motor equipped with a ball screw. Some problems of these electric driving systems are, they are excessively large-sized with complex machinery and expensive, as seen in their application to, for example, NC-machine. In order to solve these problems, the first purpose of this thesis is to develop a precision driving system actuated by a hydraulic cylinder. The hydraulic driving system consists of a cylinder and four ON/OFF solenoid valves. The second purpose is to practice the ability test of turning, drilling and milling by applying the developed driving system to a machine tool. The cutting results were compared with the ones obtained by conventional NC-machines. It was confirmed that the developed machine has a superior ability than the conventional ones.
This thesis treats block-oriented modeling, approximate linearization and control of a nonlinear pneumatic actuator system.
A block-oriented approximate feedback linearization method is proposed. For feedback linearization of a block-oriented nonlinear system, the only requirement is that some specific blocks are invertible or approximately invertible so that the manipulating calculation based on the block-units can be implemented. In this linearization, each nonlinear block can be compensated in its specific way and many nonlinearity compensation techniques can be directly incorporated.
The block-oriented approximate feedback linearization for a rodless pneumatic cylinder motion control system is demonstrated. For the linearized system, high performance force, position and tracking control can be easily achieved with just some simple linear control algorithms. The steady-state positioning error is shown to be less than the sensor's 5 µm resolution in almost full cylinder operation range and in the presence of high friction (about 15-20% of the maximum cylinder force).
A step-by-step and interactive nonlinear modeling approach for the pneumatic servo system is presented. Further, a robust re-design method, focusing on the characterization of the linearization residuals, is presented.
This dissertation focuses on innovation methods for inverter controlled hydraulic elevator system applying pressure accumulator as the "pressure-energy transformer". The required power supply and running energy can be reduced remarkably when accumulators storing and releasing pressure oil. Then the key items about hydraulic speed control system based on the EN81-2 and other hydraulic elevator standards are investigated in detail. At last, the energy-saving comparison research between the project and other hydraulic systems is completed.
In summary, the dissertation provides an innovation energy-saving hydraulic elevator control system which can reduce the required power supply remarkably. The methods of applying pressure accumulator as the pressure-energy transformer is first presented. All the research results of this thesis are not only one of the future of hydraulic elevators, but also a good chance to be applied in other vertically moved mechanism.
Aiming at the control performance problems of convergence, learning ability, robustness, and stability, this dissertation makes every effort to explore the new CMAC (Cerebellar Model Articulation Controller) neural network model structure and its control methods, with an emphasis on the analysis of the control performance. The following contents are included:
Based on the systematic research into fuzzy CMAC neural network, a general fuzzy CMAC (GFCMAC) neural network model is proposed. The expressing mode in the input space, the mapping laws in the receptive fields, and the learning algorithm are developed. The learning convergence of the GFCMAC is proved. The simulation results on the typical nonlinear functions demonstrate that the GFCMAC is not only effective, but also has a better learning performance than that of the traditional CMAC and faster convergence speed than that of the typical general basis function CMAC neural network.
To overcome the defects of the traditional basic control strategy of the CMAC neural network, the basic control strategy based on GFCMAC is first proposed. The input variables and the dimensions are determined, with the basic learning method developed. The results of simulations and experiments demonstrate that the new strategy can effectively avoid the influence of accumulative errors which can destabilize a system when tracking continuous time-varying signals, with a better tracking property obtained. Moreover, the method has a high learning rate that is important to online learning.
Based on the research into the basic control strategy, a new GFCMAC robust adaptive control strategy is proposed. The control structure and law, the input dimensions, and the weight update laws of the GFCMAC neural networks are designed. The closed-loop stability is proved based on Lyapunov stability criterion, with the update laws of the controller parameters obtained. The results of simulations and real-time control experiments demonstrate that the proposed control method is insensitive to time-varying external disturbances and uncertainties of the system parameters, having a good tracking property and strong robustness.
A novel GFCMAC reinforcement learning (RL) control method is proposed by the integration of the GFCMAC into the RL control structure, with an emphasis on the simplification and improvement in continuous time-varying signal tracking performance. The simulation and real-time control results demonstrate that the control strategy is not only effective, but also has high control accuracy. Moreover, it is easy to adjust and optimize the parameters of the GFCMAC-RL controller due to its simplicity.
A novel method of designing adaptive sliding mode equivalent control law based on the GFCMAC is proposed to attenuate the level of the chattering control input in traditional variable structure control (VSC) method, and then the tracking control performance is improved. The closed-loop stability is proved and the requirements are discussed based on the Lyapunov stability criterion. The simulation and real-time control results demonstrate that the controller is not only effective, but also insensitive to the initiative state of the system, having a good tracking control property and strong robustness. Moreover, the controller is easy to design due to the learning ability of the GFCMAC.
The proposed control methods in the dissertation are also applicable to other mechatronic control systems.
A french company (aerospace equipment supplier) plans to set up a PhD work in collaboration with the INSA Département de Génie Mécanique as soon as possible (after the summer holidays). The Phd will take 3 years to be spent most of the time in Paris. Some period will have to be done in the INSA Département de Génie Mécanique in Toulouse. The keyword of the work are: modelling, simulation, control, electrotechnics, mechanical engineering. The student will have an engineer position in the company to perform his (her) PhD work and will be paid around 10000 FF (1524 Euros) net per month.
Contact: Jean-Charles Maré, Email: jean-charles.mare@insa-tlse.fr
by: J. S. Cundiff
560 pages
Publisher: CRC Press
ISBN 0849309247
Engineers need to not only understand the basics of how fluid power components work, but they must also be able to design these components into systems and analyze or model fluid power systems and circuits. There has long been a need for a comprehensive text on fluid power systems written from an engineering perspective. Fluid Power Circuits and Controls: Fundamentals and Applications fills that need. Written for specifically for engineers rather than technicians, it builds the foundation for modeling, analyzing, and designing fluid power systems. With a unique approach that encourages readers to think of the collection of components as a system, the author illustrates each concept with a circuit diagram, and as each component is discussed, immediately places it in a circuit and analyzes its performance. Covering all aspects of the industry, this book:
The text is richly illustrated, filled with fully worked example, problems, and reinforced with exercises in each chapter. Fluid Power Circuits and Controls offers valuable design experience and the background its readers need to confidently.
Edited by: S. C. Li
500 pages
Publisher: Imperial College Press
ISBN 1-86094-257-1
An exploration of cavitation and its effects in turbines and pumps. After introducing cavitation and its relation with hydraulic machines, the invited international contributors review in detail relevant cavitation subjects from fundamental phenomena to various problems and solution measures in hydraulic machines. The authors are recognized experts in their fields.
by: P. Noskievic
276 pages
Publisher: MONTANEX Corp.
ISBN 80-7225-030-2
Language: Czech
The book System Modeling and Identification treats the methods of the building of the mathematical models of dynamic systems using the mathematical-physical modeling and the experimental identification. The application of the described methods is illustrated on the examples solved using the program MATLAB - Simulink. The book is written in Czech.
The book is based on the author's experience from the teaching at the VSB - Technical University Ostrava, Faculty of Mechanical Engineering, Department of Control Systems and Instrumentation and his experience from the modeling and simulation of the industrial problems, especially from the modeling of the electrohydraulic drives and their technological applications.
The introductory chapter deals with the basic terminology of the system theory, mathematical and physical modeling. The principles and application of the modeling and simulation are illustrated in the case study.
The mathematical-physical modeling of different systems is the topic of the following five chapters. The basic rules of the physical modeling of the real systems of different kind - mechanical systems, electrical systems, hydraulic systems, pneumatic systems and thermal systems - are given in the separate chapters. The derived mathematical models are in the form of the differential equations, transfer functions and block diagrams. The topic modeling is summarized in the seventh chapter that deals with the physical analogy.
The methods of the experimental identification are the topic of the second part of the book. The used mathematical models and input signals are shown. The characteristics of the basic linear dynamic systems are described. The methods used the deterministic input signals are explained - how to approximate the step and impulse response of the system, how to build the step response from the impulse or ramp response. The methods for parametrization of the impulse and step response are described in the next chapter.
The statistic methods of the system identification are the topic of the next three chapters. The basic statistic terms, the correlation methods and random input signals are explained.
The system identification using the discrete model parametrization is described in the next chapter. The construction of the process models AR, ARMA and dynamic systems models ARX, ARMAX, Output Error and Box-Jenkins is explained. The parametrization using the least squares method and instrumental variable methods and their recursive form are derived.
The identification of the systems operating in the closed loop is the topic of the last chapter. The direct and indirect methods with and without test signal are shown.
The printed scripts of the MATLAB commands and simulation models of the program Simulink could help with the study of the described methods. They could be also very helpful to the readers tasks solution.
The book is intended for the students and engineers that are interested in the modeling and identification of the dynamic systems. It gives the basic knowledge from the system modeling and identification. The explaining of the different identification methods could be very helpful to the application of the methods implemented in the Identification Toolbox and Signal Processing Toolbox of the MATLAB - Simulink.