A basic building block of many embedded control applications, theelectro-hydraulic servo system is used in a wide range of environments.These include manufacturing systems, materials test machines, activesuspension systems, mining machinery, fatigue testing, flightsimulation, paper machines, ships and electromagnetic marineengineering, injection moulding machines, robotics, and steel andaluminum mill equipment.
Hydraulic systems are also common in aircraft, where their highpower-to-weight ratio and precise control makes them an ideal choicefor actuation of flight surfaces.
Hydraulic actuators are characterized by their ability to impartlarge forces at high speeds and are used in many industrial motionsystems. In applications where good dynamic performance is important itis common to contain the actuator in a servo loop comprising a feedbacktransducer and electronic controller. The majority of electronicservo-controllers used in these systems are analog basedimplementations of the well-known PID(Proportional-Integrative-Derivative) type.
The most basic and obvious requirement of such systems is that thedevice must be able to compute the control algorithm swiftly enough tokeep up with the real-time demands of the system. In many cases, suchas with the simple PID controller, the control task is relativelysimple. The vast majority of electronic closed loop controllers arebased on simple analog circuit designs offering robust, low costimplementations of the well known PID control strategy.
Moving beyond MCU-based PIDcontrol
While this approach works well in systems with simple topology andlimited bandwidth, in many cases the simple MCU-based PIDcontroller is not enough. To implement more complex controlstrategies where it is often necessary to perform additional processortasks, more CPU bandwidth is required. Here it is desirable to select aprocessor which is optimized to perform real-time computations. Amulti-bus architecture and rich instruction set makes DSPs well suitedto executing demanding real-time control algorithms.
The use of digital signal processor based controllers in suchapplications is sometimes avoided because it is assumed that the shiftfrom the well understood MCU-based PID methodologies to one based onDSP will require learning new programming languages, as well as dealingwith difficulties in testing and de-bugging the code.
However, recent developments in simulation software such make itpossible to bypass such difficulties and automatically generateoptimized source code which may be compiled and run on the targetprocessor directly from the simulation environment.
This “hardware-in-the-loop“approach enables the control algorithms generated by the model to beexecuted on a real target processor during simulation, and increasesthe level of design confidence. The ease with which control algorithmscan be created and modified in this manner can save months ofdevelopment time and leads to earlier error detection compared withtraditional hand coding methods.
In this series of articles we will cover the basics ofelectro-hydraulic servo systems, pointing out where DSP-based controlloops would be useful, and describe the mathematical models for thevarious plant elements that can be developed using Simulink. Alsodescribed here is a case study of a hydraulic control system, fittingreal data to the model to validate its behavior.
The processor used to provide the basic control is the TMS320C28x(Figure 1, below ), with a 32bit fixed point DSP core. To optimize it for such applications, severallibraries have been made available for the platform, including anextensive library of optimized control algorithms.
|Figure1. TI DSP320C28X block diagram|
The basics of electro-hydraulicservo control
Although electrical motors are sometimes used in many of servo controlapplications, motion control systems requiring either very high forceor wide bandwidth are often addressed more efficiently withelectro-hydraulic rather than electromagnetic means. In general,applications with bandwidths of greater than about 20 Hz or controlpower greater than about 15 kW, may be regarded as suitable forservo-hydraulic techniques.
Apart from the ability to deliver higher forces at fast speeds,servo-hydraulic systems offer several other benefits over theirelectrical counterparts. For example, hydraulic systems aremechanically “stiffer”, resulting in higher machine frame resonantfrequencies for a given power level, higher loop gain and improveddynamic performance. They also have the important benefit of beingself-cooled since the driving fluid effectively acts as a coolingmedium carrying heat away from the actuator and flow controlcomponents. Unfortunately hydraulic systems also exhibit severalinherent non-linear effects which can complicate the control problem.
Typical Hydraulic System . Atypical position controlled hydraulic system consists of a powersupply, flow control valve, linear actuator, displacement transducer,and electronic servo-controller. The servo controller compares thesignal from the feedback displacement transducer with an input demandto determine the position error, and produces a command signal to drivethe flow control valve. The control valve adjusts the flow ofpressurized oil to move the actuator until the desired position isattained: a condition indicated by the error signal falling to zero. Aforce controlled hydraulic system operates in a similar way, exceptthat the oil flow is adjusted to achieve an output force, measured by asuitable transducer.
As shown in Figure 2, below ,all hydraulic servo systems contain at their core six basic buildingblocks: the hydraulic power supply, the flow control valve, the linerhydraulic actuator, displacement transducer, and servo controller.
|Figure2. Block Diagram of a Position Controlled Hydraulic Servo System|
Hydraulic Power supply. Systemoil pressure depends on various factors. Low pressure means lessleakage, but physically larger components are required to develop agiven force. High pressure systems suffer from more leakage, buthave better dynamic performance and are both smaller and lighter:significant advantages in mobile and aircraft applications. In manyhigh performance systems 3,000 psi (approximately 210 bar) is astandard choice of system pressure.
Oil is drawn from a reservoir (tank) into a rotary vane or pistonpump, driven at constant speed by an electric motor. The oil is drivenat constant flow rate into an adjustable pressure relief valve, whichregulates system pressure by allowing excess oil to return to thereservoir once a pre-defined pressurethreshold has been reached.
Pressurized hydraulic oil is carried to the servo-valve through asystem of rigid or flexible piping, possibly fitted with electricallyoperated shut-off valves to control hydraulic start-up and shut-downsequences. Oil is returned from the valve to the tank through a lowpressure return pipe, which is often fitted with an in-line heatexchanger for temperature regulation of the oil.
Flow control valve. Theelectro-hydraulic flow control valve acts as a high gain electrical tohydraulic transducer, the input to which is an electrical voltage orcurrent, and the output a variable flow of oil. The valve consists of aspoolwith lands machined into it, moving within a cylindrical sleeve. Thelands are aligned with apertures cut in the sleeve such that movementof the spool progressively changes the exposed aperture size and altersdifferential oil flow between two control ports.
|Figure3: Diagram of Three and Four-way Flow Control Valve Spool|
Figure 3 above shows thespool configuration of a typical “3-4” flow control valve. The portsare labeled P (pressure), T (tank), and A and B (loadcontrol ports ). The spoolis shown displaced a small distance (x v ) as a result ofa command force applied to one end, and arrows at each port indicatethe direction of fluid flow which results.
With no command force applied (F v =0 ),the spool is centralized and all ports are closed off by the landsresulting in no load flow. In the context of hydraulic servo-systems,flow control valves fall broadly into two main categories: proportionalvalves and servo-valves.
Proportional valves use direct actuation of the spool from anelectrical solenoid or torque motor, whereas servo-valves use at leastone intermediate hydraulic amplifier stage between the electricaltorque motor and the spool.
A major advantage of proportional valves is that they are largelyunaffected by changes in supply pressure and oil viscosity. However,the relatively large armature mass and large time constant associatedwith the coil means that these valves generally have poorer dynamicperformance compared with servo-valves ofequivalent flow characteristics.
In recent years, “servo-proportional” valves have begun to appearwithshorter spool displacements and lighter spools, giving dynamicperformance which approaches that of true servo-valves but at a muchlower cost.
The basic servo-valve produces a control flow proportional to inputcurrent for a constant load. While the dynamic performance of aservo-valve is influenced somewhat by operating conditions (supplypressure, input signal level, fluid and ambient temperature and so on)a major advantage is that load dynamics do not affect stability, unlikesingle stage proportional valves.
Servo valves usually have superior dynamic response, although theirclose internal machining tolerances make them relatively expensive andsusceptible to contamination of the hydraulic fluid.
Two stage servo-valves may be further divided into nozzle-flapperand jet pipe types. Both use a similar design of electromagnetic torquemotor, but the hydraulic amplifier circuits are radically different.Nozzle-flapper type servo-valves are currently by far the most commonin high performance servo applications and the description whichfollows is based on this type of valve.
|Figure4. Cross Section of Nozzle-flapper Type Servo-valve (illustrationcourtesy of Moog|
A cross sectional view of a typical nozzle-flapper type servo-valveis shown in Figure 4 above .High pressure hydraulic oil is supplied atthe inlet pressure port (P) ,and a low pressure return line to the oilreservoir is connected to the tank port (T) . The two hydraulic controlports (A and B) carry thecontrol oil flows to and from the loadactuator.
Linear Hydraulic Actuator. Ahydraulic actuator is a device which converts hydraulic energy intomechanical force or motion. Actuators may be divided into those withlinear movement (sometimes called rams, cylinders or jacks), and thosewith rotary movement (rotary actuators and motors).
Linear actuators may be further sub-divided into those in whichhydraulic pressure is applied to one side of the piston only (singleacting) and are capable of movement only in one direction, and those inwhich pressure is applied to both sides of the piston (doubleacting) and are therefore capable of controlled movement in bothdirections.
Linear actuators may also be classified as single-ended, in whichthe piston has an extension rod on one end only, or the double-endedtype which have rods on both ends. Single-ended actuators are useful inspace constrained applications, but unequal areas on each side of thepiston results in asymmetrical flow gain which can complicate thecontrol problem.
Double-ended actuators have the advantage that they naturallyproduce equal force and speed in both directions, and for this reasonare sometimes called symmetric or synchronizing cylinders. Hydraulicmotors are a separate class of actuator, in which the speed anddirection of a rotating output shaft is regulated by the flow controlvalve.
|Figure5. Linear Servo-Hydraulic Actuator Assemblies (illustrationcourtesy of Moog)|
The description which follows is based on a linear, double-acting,double-ended actuator, a type used in many industrial applications. Across section of such an actuator is shown in Figure 6. The actuatorconsists of a rod and central annulus, and incorporates low frictionseals fitted to the piston annulus and ateach of the cylinder end caps to minimize leakage. Control ports aredrilled into each end of the cylinder to allow hydraulic fluid to flowin and out of the two chambers.
|Figure6. Cross-sectional Diagram of Double-ended, Double-acting LinearActuator|
The position of the piston is determined by the hydraulic fluidpressures in the chambers on either side of the central annulus, andmay be adjusted by forcing fluid into one control port while allowingit to escape from the other. In the diagram above, hydraulic fluid isshown entering control port A while escaping from port B.
This causes an increase in fluid pressure in the chamber to the leftof the piston annulus, and a decrease in pressure in the right chamber.The net pressure difference exerts a force on the active area of theannulus which moves the piston to the right as shown. Adjustment ofpiston position is therefore a matter of controlling the differentialoil flow between the two actuator control ports.
Displacement Transducer. Position transducers are usually collocated with the actuator, andoften attached directly to the piston rod. Various types of feedbacktransducer are in use, including incremental or absolute encoders,inductive linear variable differential transformer (LVDT's) and rotaryvariable differential transformer (RVDT's), linear and rotarypotentiometers, and resolvers. In industrial applications employinglinear displacement control,the LVDT is a common choice of feedback transducer due to its accuracyand robustness.
The transducer is usually selected such that its bandwidth is tentimes or so higher than that of the servo-valve and actuator, and isoften omitted from a first analysis of the system.
|Figure7. Block Diagram of the Analogue Controller|
Servo Controller. In the Figure 7 above of the layout of theservo-controller block, the error amplifier continuously monitors theinput reference signal (u r ) and compares it against theactuator position (U p ) measured by a displacementtransducer to yield an error signal (U e ).
Ue = ur – up
The error is manipulated by the servo controller according to apre-defined control law to generate a command signal (u v )to drive the hydraulic flow control valve. Most conventionalelectro-hydraulic servo-systems use a PID form of control, occasionallyenhanced with velocity feedback. The processing ofthe error signal in such a controller is a function of theproportional, integral, and derivative gain compensation settingsaccording to the control law:
where K p , K i , and K d arethe PID constants, u e is the error signal and u v is the controller output.
With a clear idea of the basic electro-mechanical building blocksand their relationship to one another, we are now ready to develop thebasic mathematical models of the hydraulic components and make use ofthe Simulink modellingsoftwarefrom TheMathworks Inc.
This will bethe subject of Part 2 in thisseries: “Using modelling tools tosimplify hydraulic PID system design. ”
Richard Poleyis Field Application Engineer at TexasInstrumentswith focus on digital control systems .
1 . Richard Poley, DSP Controlof Electro-Hydraulic Servo Actuators, TIApplication Report SPRAA76
2. Herbert E. Merritt,Hydraulic Control Systems, Wiley, 1967
3. M.Jelali and A.Kroll,Hydraulic Servo Systems – Modelling,Identification & Control, Springer, 2003
4. Electrohydraulic Valves ATechnical Look, Moog Technical Paper
5. Moog 760 Series Servovalves,product datasheet
6. R.H.Maskrey and W.J.Thayer,A Brief History of ElectrohydraulicServomechanisms, Moog Technical Bulletin 141, June 1978
7 . T. P. Neal, PerformanceEstimation for Electrohydraulic ControlSystems, Moog Technical Bulletin 126, November 1974
8. W.J.Thayer, TransferFunctions for Moog Servovalves, Moog TechnicalBulletin 103, January 1965
9. J.C.Jones, Developments inDesign of Electrohydraulic ControlValves, Moog Technical Paper, November 1997
10. D.C.Clarke, Selection &Performance Criteria forElectrohydraulic Servovalves, October 1969
11 . M.D'Amore andG.Pellegrinetti, Dissecting High-PerformanceElectrohydraulic Valves, Machine Design,April 2001
12. D.DeRose, The ExpandingProportional and Servo Valve Marketplace, Fluid PowerJournal, March/April 2003
13. D.DeRose, Proportional andServo Valve Technology, Fluid PowerJournal, March/April 2003
14. D. Caputo, Digital MotionControl for Position and Force Loops NFPATechnicalPaper I96-11.1, April1996
15. B.C.Kuo and F.Golnaraghi,Automatic Control Systems, Wiley, 2003
16. J.Schwarzenbach andK.F.Gill, System Modelling & Control,Edward Arnold, 1992
17. J.B. Dabney and T.L.Harman, Mastering Simulink, Pearson PrenticeHall, 2004
18. Digital ControlApplications With The TMS320 Family SelectedApplication Notes (SPRA019)
19. Implementation of PID andDeadbeatControllers with the TMS320Family ( SPRA083)
20. Gene F. Franklin, J. DavidPowell & Michael L. Workman, DigitalControl of Dynamic Systems,Addison-Wesley, 1998
21 .TMS320F2812 Data Manual(SPRS174)
22. Signal Conditioning an LVDTusing a TMS320F2812 DSP (SPRA946)