Elmqvist, H., Cellier, F.E., and Otter, M., 'Object--oriented modeling of hybrid systems', in Proceedings ESS'93, European Simulation Symposium, Delft, The Netherlands, October 1993, xxxi-xli.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....design languages have been developed over the past decades. Examples include Simulink with the integration of Stateflow [17] analog and mixed signal extensions of hardware description languages such as VHDLAMS and Verilog AMS [5] object oriented modeling frameworks such as Omola[26] and Modelica [14][30] and the hybrid concurrent constraint language HCC [16] These approaches typically assume a unified model, such as a continuous time model with discontinuities, to capture semantically different components, but lack information hiding to help scale up designs. All modeling details have to be ....

H. Elmqvist, F.E. Cellier, and M. Otter, "Object-Oriented Modeling of Hybrid Systems," Proceedings of ESS'93, SCS European Simulation Symposium, Delft, The Netherlands, pp.xxxi-xli, 1993.

....recently augmented the continuous modeling with state machine based modeling of discrete control. Modeling Languages for Hybrid Systems: Tobenefitfrom object oriented design, several languages that support objectoriented modeling of complex dynamical systems have been proposed. Omola [19] Dymola [20], and Modelica [21] providenoncausalmodels; that is, there is nonotionof causality in theequationsinthemodels.Thosethreehavebeenusedmostly for describing physical objects, whereas SHIFT [22] is more like a programming language and has been used extensively to specify automated vehicle highway ....

H. Elmqvist, F. E. Cellier, and M. Otter, "Object-oriented modeling of hybrid systems," in Proc. Eur. Simulation Symp., 1993, pp. 31--41.

....augmented the continuous modeling with state machine based modeling of discrete control. Modeling languages for hybrid systems. To benefit from object oriented design, several languages that support object oriented modeling of complex dynamical sys tems have been proposed. Omola [19] Dymola [20], and Modelica [21] provide non causal models; that is, there is no notion of causality in the equations in the models. Those three have been used mostly for describing physical objects, whereas SHIFT [22] is more like a programming language and has been used extensively to specify automated ....

H. Elmqvist, F. E. Cellier, and M. Otter. Object-oriented mod- 14 eling of hybrid systems. In Proceedings of European Simulation Symposium, pages 31-41, 1993.

....language support for modeling hybrid systems to help people build complex hybrid systems for years. With an introduc tion of object oriented design paradigm [71] 13] 65] re searchers developed object oriented modeling languages for hybrid systems. Among those are Omola [56] Dymola [30], Shift [29] and Modelica [35] Those languages, however, do not provide solid formal semantics required to reason about the model of hybrid systems. In the formal verification community, formal analysis of hybrid systems has become an active research area and some tools are used successfully in ....

H. Elmqvist, F. E. Cellier, and M. Otter. Object-oriented modeling of hybrid systems. In ESS 93: European Simulation Symposium, pages 25-28, 1993.

....specified in one discrete block and one continuous block of differential equations, respectively. This model structure leads to complex interfaces of low transparency and does not support the reuse of model components. To overcome these shortcomings object oriented modeling languages like Dymola (Elmqvist et al. 1993) or Omola (Andersson 1990) were proposed. These approaches showed their suitability especially for modeling of electric circuits and mechatronic systems, but they don t a have formal semantics. Another recent approach for modeling hybrid systems was introduced by Le Bail et al. 1991) In their ....

Elmqvist, H.; F. E. Cellier; and M. Otter. 1993. Object-Oriented Modeling of Hybrid Systems. In Proceedings of the European Simulation Symposium --- ESS '93, Delft, The Netherlands, Oct. 25--28, pages 31--41.

....environments are often not able to interact, not even models can be exchanged. The Modelica design group, which started as a technical committee within EUROSIM in 1996, attempted to unify different successful concepts and to introduce a common syntax and semantic. The Modelica language [1] [2] is intended to become a neutral exchange format for model representation, but it may be simulated by commercially available simulators (e.g. Dymola [9] It shall be applicable to multiple modelling formalisms (ordinary differential equations, differentialalgebraic equations, finite state ....

Elmqvist, H.; Cellier, F.; Otter, M.: Object-oriented modeling of hybrid systems. Proc. European Simulation Symposium, ESS `93, 1993.

....detailed models are required, may we have to consider phenomena, such as thermal interactions, that are distributed in nature. There exist several modeling languages that can describe port based models. These languages are based on object oriented principles, and include Dymola (Dynasim AB 1999, Elmqvist et al. 1993), OMOLA (Anderson 1994) NMF (Sahlin 1996) and more recently Modelica (Elmqvist 1998) and VHDL AMS (IEEE 1999) Although these modeling languages can represent port based objects, they do not permit the model structure to be easily modified. Instead, only mechanisms for parameter ....

Elmqvist H., F.E. Cellier, and M. Otter. 1993. Object-oriented modeling of hybrid systems. In Proceedings of European Simulation Symposium. Delft, The Netherlands.

....approach, we need modeling tools that capture system behavior across energy domains. In recent years, a number of modeling languages have emerged that capture mathematical models of mechatronic systems. These languages are based on object oriented principles, and include Dymola (Dynasim AB 1999, Elmqvist et al. 1993), OMOLA (Ander son 1994) NMF (Sahlin 1996) and more recently Modelica (Elmqvist 1998) and VHDL AMS (IEEE 1999) Although these modeling languages are object oriented in nature, they do not permit the model structure to be easily modified. Instead, only mechanisms for parameter ....

H. Elmqvist, F. E. Cellier, and M. Otter, "Object-oriented modeling of hybrid systems," presented at European Simulation Symposium, Delft, The Netherlands, October 1993.

....are required. In recipe driven batch processes, complex logical functions must be modelled but simplified continuous dynamics may be sufficient. For batch distillation and for cyclic unit operations, complex continuous models are often necessary but the switching logic is rather simple. DYMOLA (Elmqvist et al. 1993) is a software package which allows modular modeling of dynamical systems including abrupt changes of the system structure and, e.g. hysteresis and stick slip effects based on bond graph models. Its strength is the symbolic manipulation of modularly defined equations to produce a minimal ....

Elmqvist, H., F.E. Cellier, and M. Otter (1993). Object-oriented modeling of hybrid systems. Proc. European Simulation Symposium, Delft.

....are required. In recipe driven batch processes, complex logical functions must be modelled but simplified continuous dynamics may be sufficient. For batch distillation and for cyclic unit operations, complex continuous models are often necessary but the switching logic is rather simple. DYMOLA (Elmqvist et al. 1993) is a software package which allows modular modeling of dynamical systems including abrupt changes of the system structure and, e.g. hysteresis and stickslip effects based on bond graph models. Its strength is the symbolic manipulation of modularly defined equations to produce a minimal ....

Elmqvist, H., F.E. Cellier, and M. Otter (1993). Object-oriented modeling of hybrid systems. Proc. European Simulation Symposium, Delft.

....the new state, reinitialization. Mathematical difficulties: Chattering, sliding motion, Filippov solutions. Couplings to modelling. 1996 01 25 13:51 4 Examples: Stick slip friction, relays, backlash, tank with a barrier, diodes, Petri nets, Grafcet. Reading: Chapters 5 6 in [3] and the papers [16, 29, 45]. Additional reading: The textbooks [22, 48, 18] Research Issues Model validation, techniques, formal specification of assumptions and validity ranges. Extraction of system properties from a model. Model simplification. 3. Course Plan The course will be given as a series of 10 15 lectures ....

H. Elmqvist, F. E. Cellier, and M. Otter. "Object-oriented modeling of hybrid systems." In Proceedings of the European Simulation Symposium, ESS'93, pp. 31--41, October 1993.

....too much during modelling. Another advantage is that the resulting simulation model is of less or equal numerical complexity than when using general assignment statement techniques. The modelling approach used here actually is object oriented physical systems modelling, which is currently often (Elmqvist et al. 1993; Andersson 1994; Mattson et al. 1997; Astrom et al. 1998) 20 Sim models are declarative, hierarchically structured, and encapsulation is fully supported. Furthermore, the notion of definition and use of models are distinguished (i.e. the class concept and instantiation) Note that, since bond ....

Elmqvist, H., Cellier, F.E. and Otter, M. (1993).Object-oriented modeling of hybrid systems, Proc of European Simulation Symposium, Delft, Netherlands, 31-41.

....to be a very helpful structuring aid. Mathematical representations of physical components are set up in Smile by writing down the equations describing them. Integrating equation based descriptions in an object oriented language is an unusual and promising approach (for related approaches cf. [3, 7, 8]) An equation in Smile describes the interdependence of several variables belonging to an object. Note that this constraint implies that an equation is never influenced from outside the object, thus enforcing a high degree of locality. In order to be able to construct a single set of equations ....

H. Elmqvist, F. E. Cellier, and M. Otter. Object-oriented modeling of hybrid systems. In European Simulation Symposium (ESS'93), 1993.

....much in the same way as for classical bond graphs. The code is simulated until a transition condition is met, the new mode is entered, causality propagated, analysed and simulated. This is repeated until the end time for simulation is reached. Simulation of hybrid systems is also discussed in [1, 4]. 6 Conclusions A switched bond graph is a compact graphical representation of a mode switching physical system which we want to model as a hybrid system. The representation contains all information needed for automatic generation of simulation code. Hence switched bond graphs relieves the ....

Hilding Elmqvist, Francois E. Cellier, and Martin Otter. Object-oriented modeling of hybrid systems. In Dymola - Dynamic Modeling Language, Selected Publications. Dynasim, 1994.

....many papers have covered the problem of modelling and simulation according to the hybrid approach. A majority of these papers deals with mathematical and computational issues such as the mathematical representation, the interpretation, the analysis and the practical execution of hybrid systems [6, 4, 5, 1, 14, 2]. In our work, which started in 1990, we have adopted a slightly different view. Instead of emphasising the hardcore computational matters, we have emphasised the conceptual side [12, 11, 15, 13, 8] of the modelling process. As expected, such an approach raises much debate since simulation is ....

H. Elmqvist, F.E. Cellier, and Martin Otter. Object-oriented modeling of hybrid systems. pages xxxi--xli. SCS, 1993.

....is used, the simulation tends to creep towards the discontinuity. This is the result of many rejected integration steps, since the step size prediction of the method fails at the discontinuity (Wijbrans et.al. 1992) Other research was done on combined systems simulation by Kettenis (1994) Elmqvist et al. 1993), Birta et al. 1985) Cellier (1979) Several simulation languages and programs exist that are capable of simulating combined systems, e.g. ACSL (Mitchell and Gauthier, 1993) However, to obtain efficient simulation code, one has to classify by hand the equations manually into the different ....

Elmqvist, H., F.E. Cellier, M. Otter (1993) Object oriented modeling of hybrid systems, Proc ESS 1993, pp xxxi-xli, SCS.

....1 Introduction Hybrid system is a term used in many different areas to describe system combining different properties. In this context, hybrid systems are systems that are both continuous and discrete to their nature [3] This paper deals with the problem of modeling and simulating hybrid systems [2, 6, 4]. More particularly, it deals with initializations of the continuous state variables after a discrete state change. The tool used for describing the hybrid systems is switched bond graphs. It is a tool for making graphical physical models of hybrid systems. In addition to the principle of energy ....

H. Elmqvist, F.E. Cellier, and M. Otter. Object-oriented modeling of hybrid systems. pages xxxi--xli. SCS, 1993.

....in this paper are not new and can be found in other simulation systems, too. For example the separation of model and experiment is realized in Simplex II [2] the combination of an equation based and object oriented approach can also be found in Simplex II, as well as in ObjectMath [8] and DYMOLA [1]. Still, the application of software engineering methods, techniques and tools is not common practice in the development of simulation software. To what extent the core of the Smile simulation language and the runtime support system suits the minimal basis of the family concept needs to be ....

H. Elmqvist, F. E. Cellier, and M. Otter. Object-oriented modeling of hybrid systems. In European Simulation Symposium (ESS'93), 1993.

.... yet simple, language constructs do define e.g. the value of the boolean matrix L as a function of other variables (e.g. L ii becomes one, if the generalized coordinate q i vanishes) Dymola converts such descriptions automatically into appropriate state or time events in the generated code, see [9] for details. In [26] an example of a robot with 6 dof is given, where Coulomb friction is present in every joint. Here, the above equations are used and matrices L are defined as functions of q; in accordance to the Coulomb friction model. Class Joint allows a rather general description of ....

.... chemical reaction systems [3] electronic circuits [14] and a detailed robot model [26] Furthermore, Dymola has sophisticated language on an Apollo workstation with a Motorola 68040 processor elements for the definition and handling of discrete event systems based on time and state events [9]. It is therefore easy to model e.g. sampled data systems, discontinuous elements, interconnected friction elements, or impact. Appendix: Derivation of Basic Mechanical Equations In this section, equations (4) are derived. Equations (4a 4e) are directly given by their definition. Equation (4f) ....

Elmqvist, H., Cellier, F.E., and Otter, M., 'Object--oriented modeling of hybrid systems', in Proceedings ESS'93, European Simulation Symposium, Delft, The Netherlands, October 1993, xxxi-xli.

....attention is now given to discrete modeling features. 4S YNCHRONOUS EQUATIONS In Modelica the central property is the usage of synchronous differential, algebraic and discrete equations. The idea of using the synchronous data flow principle in the context of hybrid systems was introduced in [6]. For pure discrete event systems, the same principle is utilized in synchronous languages [8] such as SattLine [5] Lustre [9] and Signal [7] in order to arrive at safe implementations of realtime systems and for verification purposes. A typical example of a hybrid Modelica model is given in ....

....leads to a system of mixed continuous discrete equations which have to be solved at event instants. In the past, ideal switching elements have been handled by (a) using variable structure equations which are controlled by state transition diagrams to describe the switching behaviour, see e.g. [2, 6, 13], or by (b) using a complementarity formulation, see e.g. 10, 16, 17] The approach (a) has the disadvantage that the continuous part is described in a declarative way but not the part describing the switching behaviour. As a result, e.g. algorithms with better convergence properties for the ....

H. Elmqvist, F.E. Cellier, and M. Otter. Object--Oriented Modeling of Hybrid Systems. In Proceedings ESS'93, European Simulation Symposium, pages xxxi--xli, Delft, The Netherlands, 1993. Society for Computer Simulation

....modeling with true equations and the use of objectoriented constructs to facilitate reuse of modeling knowledge. There are already several modeling languages with such a support available from universities and small companies. Examples of such modeling languages are ASCEND [16, 15] Dymola[7, 6],gPROMS [2, 14] NMF[17] ObjectMath [8, 20] Omola[13, 1] SIDOPS [4] Smile [3, 12] U.L.M. 11, 10] and VHDLAMS [9, 5] There is also significant experience of using these languages in various applications. The aim of the Modelica effort is to unify the concepts of these languages in ....

H. Elmqvist, F. Cellier, and M. Otter. "Object-oriented modeling of hybrid systems." In Proceedings of European Simulation Symposium, ESS'93.TheSocietyof Computer Simulation, October 1993.

.... yet simple, language constructs do define e.g. the value of the boolean matrix L as a function of other variables (e.g. L ii becomes one, if the generalized coordinate q i vanishes) Dymola converts such descriptions automatically into appropriate state or time events in the generated code, see [9] for details. In [26] an example of a robot with 6 dof is given, where Coulomb friction is present in every joint. Here, the above equations are used and matrices L are defined as functions of q; l in accordance to the Coulomb friction model. Class Joint allows a rather general description of ....

....[14] and a detailed robot model [26] Furthermore, Dymola has sophisticated language 6 on an Apollo workstation with a Motorola 68040 processor Nonlinear Dynamics , Vol. 9, pp. 91 112, 1996. 25 elements for the definition and handling of discrete event systems based on time and state events [9]. It is therefore easy to model e.g. sampled data systems, discontinuous elements, interconnected friction elements, or impact. Appendix: Derivation of Basic Mechanical Equations In this section, equations (4) are derived. Equations (4a 4e) are directly given by their definition. Equation (4f) ....

Elmqvist, H., Cellier, F.E., and Otter, M., 'Object--oriented modeling of hybrid systems', in Proceedings ESS'93, European Simulation Symposium, Delft, The Netherlands, October 1993, xxxi-xli.

....by now that separating the modeling code into initial, derivative, and discrete sections, as done in ACSL, Omola, VHDL A and other systems, is not a good idea in the presence of state events. For the user it is nearly impossible to manually generate code that is correct in all circumstances. In [Elmqvist, et al. 1993], a satisfactory solution to the problems mentioned above is proposed for object oriented modeling systems, and the proposed solution has been implemented in Dymola. It is beyond the scope of this article to discuss all the details. In a nut shell, higher language elements are introduced that ....

Elmqvist, H., Cellier, F.E., and Otter, M. 1993. Object--oriented modeling of hybrid systems. Proc. ESS'93, European Simulation Symp., Delft, The Netherlands, pp. xxxi--xli.

....1: Bond graph representation of Sw element. The switch equation states that, at all times, either f or e are zero, depending on whether the switch is open (OpenSwitch=true) or closed (OpenSwitch=false) respectively. Dymola is able to cope with such a variable structure equation as shown in (Elmqvist et al. 1993). If the switch is open (f = 0) the Sw model computes the flow, and the causality stroke should be at the Sw element. If the switch is closed (e = 0) the Sw model computes the effort, and the causality stroke should be away from the Sw element. Thus, the causality stroke indeed changes its ....

....between which the simulation program would have to toggle whenever switching takes place in the circuit. This is, however, an unfortunate proposition since a circuit containing 10 switch elements would call for 2 10 = 1024 different models. A solution to this dilemma was finally proposed in (Elmqvist et al. 1993). A single model that works in both switch positions can easily be derived if and only if both causalities are compatible with the remainder of the bond graph, i.e. if changing the causality at the Sw element does not force any incorrect causality on sources or undesired causality on storage ....

[Article contains additional citation context not shown here]

Elmqvist, H., F.E. Cellier, and M. Otter. 1993. "Object--Oriented Modeling of Hybrid Systems." In Proc. ESS'93, European Simulation Symposium (Delft, The Netherlands, October 25--28), xxxi-- xli.

....cost efficient. This leads to the introduction of model classes. A notation to declare connection mechanisms, cuts and their semantics is also needed. Such facilities are available in the object oriented language Dymola (Elmqvist, 1995) Instantaneous equations are used to handle discrete effects (Elmqvist, et al. 1993). This paper discusses the most recent developments, such as matrix notation, multiple inheritance and generic class parameters. 2. SEMANTIC TRANSFORMATIONS Object oriented language constructs are very efficient for the modeler. It is, however, essential that numerical routines can solve the ....

....thus needs to be transformed to efficient algorithms for calculations. Graph theoretical algorithms are used for the structured analysis needed to determine causalities. Computer algebra is used to solve equations for their unknowns and performing partial evaluation at compile time, Cellier and Elmqvist, 1993). Interfacing to numerical routines is described in (Otter and Elmqvist, 1995) An new efficient mixed symbolic numerical approach for solving differentialalgebraic equations is presented in (Elmqvist, et al. 1995) 3. MATRIX EQUATIONS In mechanical modeling, matrices are essential to represent ....

[Article contains additional citation context not shown here]

Elmqvist, H., F.E. Cellier, and M. Otter (1993), "Object-oriented Modeling of Hybrid Systems," ESS'93, European Simulation Symposium, Delft, The Netherlands, October 25-28, 1993.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC