Documents

1430

Description
gfj
Categories
Published
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
    Session 12b6   0-7803-5643-8/99/$10.00 © 1999 IEEE November 10 - 13, 1999 San Juan, Puerto Rico 29 th  ASEE/IEEE Frontiers in Education Conference 12b6-7   Hardware-in-the-loop simulation and its application in control education Wojciech Grega Department of Automatics, University of Mining and Metallurgy 30-059 Krakow, Al.Mickiewicza 30, Poland wgr@aquarium.ia.agh.edu.pl  Abstract   This paper describes the concept of “hardware-in-the-loop” design system implemented as an experimental  part of the control engineering course. The concept of „hardware-in-the-loop” (HiL) method is to use a simulation model of the plant and the industrial standard target controller. The integrated analysis and simulation environment supports the design system. Hardware and  software components of the HiL system are presented in the  paper and an example of the laboratory exercise is described to give insight into the approach, which has been adopted.   Introduction It has long been recognised that the laboratory experiment is an important way in which abstract concepts of engineering can be related to the design problems. This is a  particulary true in control engineering education. Students find the control theory hard to understand, limited in its application and become less well motivated. Therefore, the use of laboratory exercises which illustrate the theory at work in practical situations is an important aspect of control engineering education [1],[2]. The srcinal motivation behind constructing laboratory scale models of industrial processes was to offer the students an environment for a complete engineering design, starting from modelling and simulation and ending with the experimental verification of a wide range of control strategies. A typical integrated real-time control and simulation environment contains three main parts: analysis and simulation software, a target controller and an experimental set-up. The software includes controller design/analysing tools, real-time code generators and a compiler. The target hardware can be based upon DSP techniques or other low-cost alternatives such as PC-based controllers or microcontrollers can be used. Usually the costs of the first two components are much lower in comparison with the costs of the experimental set-up. Capturing the realism of industrial control problems requires a complex installation or machinery, both costly and inflexible, in some cases potentially dangerous for the students. The concept of „hardware-in-the-loop” (HiL) method is to use a simulation model of the process and the real target hardware. The simulation model provides all the  process signals in real-time that are next converted by D/A modules and supplied to the controller as voltages. The control signals are produced by the controller and supplied via A/D converters to the simulation model. It is the  purpose of this configuration to make the hardware component behave as closely as possible to these that would  be encountered in the real system. The paper is organised as follows. In the first section the concept of the integrated real-time control and simulation environment is explained. Next, we describe hardware and software components of the HiL design system and present the application of the HiL method for control of the ventilation system. Finally, the benefits of the  proposed solution are discussed and compared with other alternatives. Computer-aided control system design and real-time control A typical development cycle of control system design and implementation is given in Figure 1.   control & monitoring real-time tasks  Design Processanalysis Matla Processsimulation Simulink  CACD tools 4 <: B code ritingcompilatioloading identification target controller    Figure 1. Typical development cycle of a control system    Session 12b6   0-7803-5643-8/99/$10.00 © 1999 IEEE November 10 - 13, 1999 San Juan, Puerto Rico 29 th  ASEE/IEEE Frontiers in Education Conference 12b6-8  The design process starts with determination of goals and modelling of the process to be controlled, and ends with experiments. Modelling is to formulate a mathematical description that characterises the input-output relationships of the process to be controlled. The model can be obtained  be applying physical laws or by identification experiments. This is followed by a design phase, which requires selection of control strategies, structures and parameter values.  Nowadays, designing methods are supported by computer-aided analysis and simulation tools [3]. Most of them aid the control system engineer with a complete software environment for the analysis and synthesis aspects of control systems. The analysis, modelling and simulation cycle is iterative. After the design process is completed the implementation phase requires translation of a control algorithm into the real-time control code . The main features of the real-time software, as distinct from other software, are that the control algorithms must be run at their scheduled sample intervals and that there exists associate software components, which interact with the sensors and actuators. This implementation phase requires from student some skills beyond the basic application of  programming languages, like interrupt handling, interaction with an external hardware, task scheduling. Finally, the control code is loaded to the target controller. The controller is then integrated with the  physical system for further tests, using physically measured inputs and generated outputs. The implementation  procedure can also be performed several times with modified parameters or structures of the control algorithm until the design goals are met. In recent years the development of computer-aided control design tools has changed the control systems design  practice significantly, both in industry and educational sectors [4], [5]. The  Integrated Real-time Control and Simulation Environment   (IRCSE) is a tool enabling the designer to perform simulations and real-time experiments in structured and simple manner. Typically, the IRCSE consists of three main parts (Figure 2): ã real-time part, referred as a real-time kernel (RTK), ã analysis and simulation software available in „on-line” mode, ã generator of real-time code. The  real-time kernel (RTK   ) , installed on the target controller, performs the interfacing operation, control algorithm and data logging. Data collected in the cyclic  buffer of the real-time kernel are transferred to the CACD environment and are available in „on-line” mode. A special Graphical User Interface  (GUI) is created for “on-line” visualisation of process data. The main advantage is the  possibility of using simulation and analysis procedures as remotely communicating programs including interactive “on-line” parameter tuning. control &communication real-time tasks  Design Processanalysis Matlab processsimulation Simulink  CACD tools : code  generatorcompilationloadingGUI on-line tasks  target controller    Figure 2. Development cycle of a control system in the  IRCSE The code generator supports rapid prototyping   of the control algorithm. The code generator produces a run-time code based on the controller designed in the CACD environment. Figure 3 shows the general concept that comes within the scope of the code generator. We wish to use simulation tools to design, test and analyse a model + Modelof the processModelof the controllerModelof the controllerInputDriverOutputDriverReal-time control algorithm    Figure 3. Controller rapid prototyping    Session 12b6   0-7803-5643-8/99/$10.00 © 1999 IEEE November 10 - 13, 1999 San Juan, Puerto Rico 29 th  ASEE/IEEE Frontiers in Education Conference 12b6-9  of a control system. Once the project is complete we can add an appropriate input/output drivers to generate an executable code directly from the block diagram. The code generator software translates the controller diagram into its equivalent code representation, adding functions that are unique to the target controller. This solution offers automation of the controller design freeing the designer from the burden of manual PLC programming. The concepts presented above are supported by a number of commercially available or srcinally developed rapid prototyping tools (see Table 1). Educational applications focus on personal computers (PCs) which, along with a related real-time extension of the Windows 95/98/NT operating systems, can provide  low cost solutions to the process control of small plants or laboratory installations.  Table 1. Examples of IRCSE & rapid prototyping software tools Product Target Platform CACD environment Visualisation tools Remarks dSpace ( dSpace-GmbH  ) DSP MATLAB/SIMULINK SIMULINK high-speed control and data analysis Paradym (  Intellution ) PLC, industrial PC Sequential function charts, function blocks FIX-Intellution industrial standard:IEC 1131-3 RT-CON (  Inteco Ltd  ) PC, PLC MATLAB/SIMULINK SIMULINK low-cost solution What makes this approach especially effective for educational applications is the general assumption that the integrated environment creates an open architecture  process automation system . It uses a standard hardware  platform, standard operating system and commercial Computer Aided Control Design Systems as engineering tools. MATLAB  ,  a popular model analysis package and SIMULINK   , object oriented dynamic system simulation  package  ,  were selected as CACD tools in the application described in this paper. They include a number of toolboxes for signal processing, system identification, and control design, both for linear and non-linear, single input - single output (SISO) and multi input-multi output (MIMO) systems. However, drivers for communication between the RTK and analysis/simulation software are not included in the standard versions. To overcome this interprocess communication functions must be developed as an extension of MATLAB software to allow data transfer  between the real-time part and the CACD. The code generator uses  Real Time Workshop  toolbox supported by RT-CON software – a group of tools meant to cover the different real–time code generation stages in the Windows95/98/NT environment. Laboratory scale models in control education The srcinal motivation behind constructing laboratory scale models is to illustrate theory at work in practical situations and to offer the students an environment for a complete engineering design, starting from modelling and simulation and ending with the experimental verification. The laboratory sessions attempts to reflect the industrial aspects of control engineering by carrying out application-oriented projects. The general assumption is to capture the realism of industrial control problems in a laboratory environment. In many cases development of the models is stimulated by industrial-commercial applications of new results. In this respect some typical areas were selected in control laboratories to be representative as the prototypes of realistic computer controlled processes from the various  branches of control system engineering: electromechanical systems [6],[7] (inverted pendulum, helicopter model, digital servo, magnetic bearing, simple robots), fluid-level control [8], thermal and chemical process control (e.g. PH-control). In most cases the technical importance of the models is immediately evident for the students allowing them to relate the theory to the physical world. However, more complex laboratory models of industrial plants are too expensive for control education. The isolated initiatives in the development of prototypes of complex laboratory installations were reported, but always supported by industry or research grants. As a matter of fact, experimental verification of control algorithms requires a computer-controlled source of the  process signals. For example, the signals can be produced  by a simulation model of the plant and supplied via D/A converters to the PLC controller. The only requirement is real-time behaviour of the simulation model. This concept is described in the next section. Hardware-in-the-loop simulation Generally, a classical control experiment requires a direct connection with the plant (Figure 4). A plant or its laboratory model delivers physically measured inputs. Generated outputs are send to the actuators: motors, valves, amplifiers, etc. This is a classical    Session 12b6   0-7803-5643-8/99/$10.00 © 1999 IEEE November 10 - 13, 1999 San Juan, Puerto Rico 29 th  ASEE/IEEE Frontiers in Education Conference 12b6-10  design approach based on experimental validation of the of the control algorithm. GE PROG +210 A/DD/A RS   PLC plant I devicebus    Figure 4. Classical control experimental set-up However, we can easily imagine a reverse situation, when the process from Figures 4 or 2 is substituted by its model and is being simulated in real time. The controller is fully implemented in this case. This approach is called hardware-in-the loop simulation . The simulated plant  provides all input signal to the real PLC controller (Figure 5). In this scenario fully programmed, industrial standard controller interfaces with real-time simulation. 1 1 1 1 1 1 1 1   1 PROGR +2410 A/DD/A RS PLC PC1PC2I/O board    Figure 5. Hardware-in-the loop method This general concept was first implemented by dSPACE GmbH company [9] resulting in a line of products  based on TMS320 DSP processors and specialised toolboxes for hardware-in-the-loop simulation. The dSPACE Real-Time Interface offers automation of the hardware-in-the-loop experiments supported by high-speed of digital  processor system. The only drawback of this configuration is the price: the code generation software is very expensive due to its small market share. The proposed system consists of an industrial-standard  programmable logic controller (PLC), the PC–type computer where the simulation model is located (PC1) and another PC used for PLC code development (PC2). The low-cost input/output board is plugged into slot of the PC1 computer. The board features 16 analog input channels, two analog output channels and 16 configurable digital I/O channels. The key to hardware-in-the-loop simulation method lies in the software. In this experimental set-up the software includes (Figure 6): ã tools for modelling and simulation (MATLAB/SIMULINK), available as a low-cost educational licence, ã library of I/O drivers: a set of pre-defined blocks, representing the functions implemented in hardware, ã code generation software for automatic building of real-time models (RTW, RT-CON software), ã user-interface for visualisation of data and on-line tuning of the parameters.   A/D WINDOWS 95/98/NT, RTW  Real –time kernel I/O board isualisation from sensors D/A o actuators  I/O driver MATLAB/ SIMULINK  communictioninterfacetunning real-timemodel  A123   B123 A123   B123    A123   B123    A123   B123    A123   B123    A123   B123    A123   B123    A123   B123    A123   B123POWE   GEFan PROGRAMMAB +24V100- PLC controller   A/DD/A DigitalI/O ½    Figure 6. HiL software and hardware architecture The major functions assigned to the particular parts of this software environment are as follows: ã development of the simulation model of the plant (MATLAB/SIMULINK). Next, the model is supplied
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x