2ND SWEDISH-ISRAELI CONTROL CONFERENCEiaac.technion.ac.il/SwIsr/SwIsr brochure.pdf · (lumped) electrical networks theory (R-L-C), by W. Cauer, R. M. Foster, O. Brune, and others

Room 217, D. Dan and Betty Kahn Bld.Technion—Israel Institute of Technology

Haifa, Israel, November 9–11, 2014

2ND SWEDISH-ISRAELICONTROL CONFERENCE

Sponsoring Organizations

IAAC — Israeli Association for Automatic ControlThe Center for Mathematical Sciences at the Technion

The Bernard M. Gordon Center for Systems Engineering at the TechnionThe Moshe Yanai Fund for the Promotion of International Conferences at the Technion

Faculty of Mechanical Engineering at the TechnionSwedish Technion Society

Organizers

Bo Bernhardsson (Lund University)Maxim Kristalny (Technion—IIT)Leonid Mirkin (Technion—IIT)

Workshop Venue

Room 217, D. Dan and Betty Kahn Bld, Faculty of Mechanical Eng.

Venue

Forchheimer

Den andra svensk-israeliska konferensen i reglerteknik הכנס השבדי־ישראלי השני לבקרהhttp://iaac.technion.ac.il/SwIsr/

http://iaac.technion.ac.il/SwIsr/

http://iaac.technion.ac.il/SwIsr/

Technical ProgramSunday, November 9, 2014

8:15 Gathering & registration

8:45 Opening

9:00 Anders Lindquist, KTH & Jiao Tong UIntermittent Observations and Hybrid Dynamics: The Reversal of the Time Direction in Stochastic Models and its Use in Optimal Estimation

9:40 Izchak Lewkowicz, Ben-Gurion U(Generalized) Positive Rational Functions — Convex Invertible Cones Overview

10:20 Coffee break

10:50 Thomas Schön, Uppsala UNonlinear System Identification Enabled via Sequential Monte Carlo

11:30 Jonas Sjöberg, Chalmers UIdentification of Wiener-Hammerstein Models and Other Nonlinear Block Models

12:10 Lunch (Forchheimer)

14:00 Alexander Medvedev, Uppsala UPointwise and Distributed Delays in Impulsive Models of Endocrine Regulation

14:40 Michael Margaliot, Tel-Aviv UThe Ribosome Flow Model: Theory and Applications

15:20 Coffee break

15:50 Thomas Gustafsson, Luleå UTowards Optimal Control of Wood Drying

16:30 Yizhar Or, TechnionDynamics and Control of (Bio-)Robotic Locomotion: Nonlinear, Nonholonomic and Hybrid Mechanical Systems

Monday, November 10, 2014


9:00 Shai Arogeti, Ben-Gurion USet-point Regulation of an Uncertain 6-DOF Magnetically Levitated Positioning Stage

9:40 Karl Johan Åström, Lund UControl Design for a MEMS Accelerometer with Tunneling Sensing

10:20 Coffee break

10:50 Bo Bernhardsson, Lund UControl of a Marine Vibrator Using Iterative Learning Control

11:30 Tal Shima, TechnionCooperative Optimal Missile Guidance Laws

Lunch, Trips & Banquet

Tuesday, November 11, 2014


9:00 Ilana Nisky, Ben-Gurion USurgical Robotics—Enhancing the Surgeon in the Loop

9:40 Maxim Kristalny, TechnionOptimal Control in Delayed Bilateral Teleoperation

10:20 Coffee break

10:50 Daniel Żelazo, TechnionUncertain Consensus Networks: Robustness and its Connection to Effective Resistance

11:30 Giacomo Como, Lund UResilient Distributed Routing in Dynamical Flow Networks

12:10 Lunch (Forchheimer)

14:00 Anders Rantzer, Lund UScalable Control of Positive Systems

14:40 Karl Henrik Johansson, KTHControl and Optimization of Cooperative Driving for Road Goods Transportation

15:20 Jack Haddad, TechnionOptimal Perimeter Control Synthesis for Two Urban Regions with Boundary Queue Dynamics

16:00 Coffee break

16:30 Karl Johan Åström, Lund UControl—A Perspective

17:30 Farewell

Den andra svensk-israeliska konferensen i reglerteknik הכנס השבדי־ישראלי השני לבקרה

12:4

0–21

:00

syst

ems

theo

rysy

stem

iden

tifica

tion

biol

ogic

al s

yste

ms

cont

rol a

pplic

atio

ns

cont

rol a

pplic

atio

nslig

htly

-dam

ped

syst

ems

tele

oper

atio

nne

twor

k co

ntro

ldi

stri

bute

d &

traf

fic c

ontr

olke

ynot

e

Abstracts | Sunday, November 9, 20149:00–9:40

Anders Lindquist (Royal Institute of Technology & Shanghai Jiao Tong University)

Title: Intermittent Observations and Hybrid Dynamics: The Reversal of the Time Direction in Stochastic Models and its Use in Optimal Estimation

Abstract: We consider data fusion for the purpose of smoothing and interpolation based on observation records with missing data. The stochastic processes are generated by linear stochastic models, possibly hybrid. The reversal of the time direction in stochastic models allows a unified framework where the estimation is effected by the fusion of the estimates of two filters running in opposite time directions. These two filters integrate the two complementing parts of the available observation record, past and future, respectively, and their estimates are weighed in using a simple Mayne-Fraser-like formula. The paper begins by establishing a connection between time reversal in stochastic systems and all-pass extensions. A particular normalization (choice of basis) between the two time-directions allows the two to share the same sample paths and simplifies the mathematics of optimal interpolation and optimal smoothing. It is seen that the framework allows treatment of hybrid stochastic models and intermittency in observations in a rather straightforward manner.

This is joint work with T. T. Georgiou.

9:40–10:20

Izchak Lewkowicz (Ben-Gurion University of the Negev)

Title: (Generalized) Positive Rational Functions—Convex Invertible Cones Overview

Abstract: Convex Invertible Cones (CICs in short) of matrices are sets which are closed under positive scaling summation and inversion (whenever the matrix is non-singular). We first relate these sets to the Lyapunov equation. Specifically, to the set of matrices sharing the same Lyapunov factor (“common Lyapunov solutions”).

Then the CIC structure is associated with the Sylvester and the Riccati equations with prescribed common solutions.

Rational positive functions, denoted by P, are analytic in the open right half of the complex plane and have positive semi-definite Hermitian part in its closure.

Physical systems are often real, so engineers typically refer to Positive Real functions. In this talk, we relate to both the real and the complex frameworks.

The study of rational positive functions has been motivated from the 1920's by (lumped) electrical networks theory (R-L-C), by W. Cauer, R. M. Foster, O. Brune, and others. From the 1960's positive functions also appeared in books on absolute stability theory, e.g. of V. Popov.

The set P forms a CIC of matrix-valued rational functions. There are several interrelations between CICs of constant matrices and of rational functions. Some will be presented. In particular, the Kalman-Yakubovich-Popov Lemma.

As time permits, Generalized Positive (mixed inertia) extensions of the above results will be presented.

We shall conclude with novel idea of applying classical synthesis techniques of R-L-C circuits to the construction of numerous feedback-loops systems.Joint work with D. Alpay, Math. Dept., Ben-Gurion University, Israel.

10:50–11:30

Thomas Schön (Uppsala University)

Title: Nonlinear System Identification Enabled via Sequential Monte Carlo

Abstract: Sequential Monte Carlo (SMC) methods are computational methods primarily used to deal with the state inference problem in nonlinear state space models. The particle filters and the particle smoothers are the most popular SMC methods. These methods open up for nonlinear system identification (both maximum likelihood and Bayesian solutions) in a systematic way. As we will see it is not a matter of directly applying the SMC algorithms, but there are several ways in which they enter as a natural part of the solution. The use of SMC for nonlinear system identification is a relatively recent development and the aim here is to first provide a brief overview of how SMC can be used in solving challenging nonlinear system identification problems by sketching both maximum likelihood and Bayesian solutions. We will then introduce a recent powerful class of algorithms collectively referred to as Particle Markov Chain Monte Carlo (PMCMC) targeting the Bayesian problem. PMCMC provides a systematic way of combining SMC and MCMC, where SMC is used to construct the high-dimensional proposal density for the MCMC sampler. The first results emerged in 2010 and since then we have witnessed a steadily increasing activity within this area.


11:30–12:10

Jonas Sjöberg (Chalmers University of Technology)

Title: Identification of Wiener-Hammerstein Models and Other Nonlinear Block Models

Abstract: Block based nonlinear models consisting of connected blocks containing either LTI transfer functions or parameterized static one-dimensional mapping are popular nonlinear models. They are are easier to analyse and understand than general nonlinear dynamic models, and they are, typically, more sparsely parameterized which gives less problems with over-fitting.

This talk focus on Wiener-Hammerstein models which consists of two LTI blocks with a static nonlinearity in-between. Many suggested methods starts with a linear approximation which then needs to be splitter into two blocks. Some examples of suggested methods in the literature, their strengths and weaknesses are given. Some new results from an approach where the linear blocks are initialised with the original liner model but with a parameterized fraction so that the two blocks in series becomes equivalent to the linear model. The fraction LTI blocks are then approximated with series expansions so that only integer delays of the data need to be handled.

14:00–14:40

Alexander Medvedev (Uppsala University)

Title: Pointwise and Distributed Delays in Impulsive Models of Endocrine Regulation

Abstract: Hormones are molecules that constitute the basis of chemical signaling in a multi-cellular organism. Many hormones are secreted in pulses and can therefore communicate information to the target cells not only through their concentration but also via amplitude and frequency of the pulse train. Thus the purpose of pulsatile endocrine control is in invoking suitable dynamical signaling patterns.

Pulsatile endocrine regulation is typically implemented by the nervous system exerting feedback control action on much slower chemical dynamics. This can be exemplified by the gonadotropin-releasing hormone (GnRH) that mediates the central control of human reproduction and is secreted in the hypothalamus. In the male, GnRH stimulates the production of luteinizing hormone (LH) in the

hypophysis that, in turn, stimulates the synthesis of testosterone (Te) in the testes. Te inhibits the production of GnRH, which phenomenon closes a negative feedback loop over the GnRH-LH-Te axis. The pulsatile secretion of GnRH is crucial to the function as constant levels of the hormone effectively suppress the production of LH instead of stimulating it.

The machinery of hybrid systems is very much suitable for mathematical description of pulsatile endocrine regulation as the chemical plant is inherently continuous while the regulatory feedback clearly exhibits discrete dynamics via a pulse-modulated feedback mechanism. The resulting closed-loop behaviors are highly nonlinear due to the nature of pulse-modulated control. Multiple delays occur in the hormone control loops due to transport phenomena. Furthermore, distributed delays are often utilized in mathematical modeling to account for the interaction between the hormone molecules and receptors.

Analysis of impulsive endocrine Te regulation models with pointwise and distributed time delays reveals a number of intriguing facts. First, it appears to be very difficult to distinguish between finite-dimensional and infinite-dimensional dynamics due to so-called finite-dimensional reducibility property of the continuous part of the loop. Second, the parameters of the frequency modulation law set an upper bound to the duration of the time delay that does not cause qualitative changes in the closed-loop system dynamics. When this bound is violated, the bifurcation scenarios of the hybrid dynamics become virtually unlimited. Third, for biologically motivated parameter values, the pointwise time delay has a stabilizing action as the periodic solutions in the closed-loop possess lower multiplicity for higher delay values.

14:40–15:20

Michael Margaliot (Tel-Aviv University)

Title: The Ribosome Flow Model: Theory and Applications

Abstract: A crucial stage in the production of proteins from the information encoded in the genes is called translation. During this stage, complex molecular machines, called ribosomes, bind to the mRNA and “read” it in a sequential manner.

In 2011, Reuveni et alii suggested a new nonlinear model for this process called the Ribosome Flow Model (RFM). The RFM can be derived as a mean-field approximation of an important model from non-equilibrium statistical physics called the Totally Asymmetric Simple Exclusion Process (TASEP).

Abstracts | Sunday, November 9, 2014


In this talk, we study the RFM using tools from systems and control theory including contraction theory, monotone systems theory, the analytic theory of continued fractions, and convex analysis. We detail several biological implications of the analysis and compare them to known experimental results.Joint work with Tamir Tuller (Tel Aviv U) and Eduardo D. Sontag (Rutgers U)

15:50–16:30

Thomas Gustafsson (Luleå University of Technology)

Title: Towards Optimal Control of Wood Drying

Abstract: Wood Drying is an energy intensive business which today is mainly controlled by precomputed schedules. The corrosive environment in a drying chamber with high humidity and temperature makes the sawmills often choose to have as few sensors as possible, usually just dry and wet bulb temperature. This presentation discusses the possibilities to estimate the moisture content and tension in the timber based on the temperature measurements. The goal is to design an optimal controller that minimizes energy consumption and drying time while the tension in the wood is kept at a sufficiently low level to avoid cracking and other defects. Results from full-scale experiments will be presented.

The work is done in collaboration with a company that supplies control systems for drying kilns.

16:30–17:10

Yizhar Or (Technion—Israel Institute of Technology)

Title: Dynamics and Control of (Bio-)Robotic Locomotion: Nonlinear, Nonholonomic and Hybrid Mechanical Systems

Abstract: Locomotion of mobile robots or moving creatures is generated by internal actuation of shape changes combined with physical interaction with the surrounding environment. The mechanics of locomotion typically gives rise to nonlinear control systems which are governed by nonholonomic constraints. In this talk, several (bio-) robotic locomotion systems will be overviewed: Swimming microorganisms and micro-robotic swimmers whose motion is governed by low-Reynolds-number hydrodynamics, in which inertial effects are negligible, can be controlled by internal motors or by actuation of external magnetic fields. Underactuated vehicles are governed by nonholonomic mechanics due to no-slip constraints of wheels, and their motion can be formulated as nonlinear control systems. The dynamics of legged locomotion involves nonsmooth transitions between different states of ground-foot contact, and can be formulated as hybrid systems. It will be shown how one can use simple models of the locomotion systems' dynamics and mechanical actuation in order to analyze their control and stabilization capabilities and optimize their performance.

Abstracts | Sunday, November 9, 2014


Abstracts | Monday, November 10, 20149:00–9:40

Shai Arogeti (Ben-Gurion University of the Negev)

Title: Set-point Regulation of an Uncertain 6-DOF Magnetically Levitated Positioning Stage

Abstract: A novel 6-DOF magnetically levitated positioning stage is presented and its set-point control problem is discussed. In particular, an iterative output feedback control scheme is proposed that is based on position measurements only. Due to an unknown payload (in terms of both, mass and location) the stage model is uncertain. A learning mechanism is proposed that is based on the well-known contraction mapping principle. Besides stability and convergence analysis, the presented approach includes some aspects related to the practical implementation of the algorithm. In particular, it is shown that: (i) the system response due to the iterative process converges to an arbitrarily small neighborhood of the desired equilibrium point, and (ii) the steady state response after each iteration is bounded inside the system traveling range. The proposed control strategy was verified experimentally.

9:40–10:20

Karl Johan Åström (Lund University)

Title: Control Design for a MEMS Accelerometer with Tunneling Sensing

Abstract: MEMS devices typically have resonances with low damping. Very precise control can be obtained by sensing position using tunneling sensors. The devices are subject to random disturbances such as, Brownian motion, Johnson-Nyquist noise, and tunneling noise. Variations in device parameters must also be accounted for. In this paper we will discuss principles for designing control systems for instruments based on MEMS devices with force feedback. The fact that the final goal is to design an instrument gives an interesting formulation of the control problem, which make it possible to separate shaping of the frequency response from attenuation of disturbance. The principles have been applied to design of a tunneling accelerometer with position sensing based on quantum tunneling. Results from laboratory experiments will be presented.

The major part of the work has been done in Professor Kimberly Turners laboratory at the University of California Santa Barbara.

10:50–11:30

Bo Bernhardsson (Lund University)

Title: Control of a Marine Vibrator Using Iterative Learning Control

Abstract: Marine vibrators, large loud speakers, are used for generating a powerful underwater acoustic signal. A receiver array is then used to process the information in echoes to make a map of the area below the bottom of the ocean and e.g. find possible areas of oil. To minimize impact on animal life, future restrictions on the spectrum content of the sound signals make it necessary to improve the linearity of the vibrators. This is a difficult engineering challenge since the vibrators are large electro-mechanical constructions with severe nonlinearities in the form of e.g. friction and backlash. Fortunately, the behavior is quite repeatable, which makes the problem a suitable candidate for iterative learning control (ILC). We will describe the successful design, implementation and experiments of frequency domain based ILC algorithms reducing the harmonic overtones to acceptable levels.

Joint work with Olof Sörnmo, Olle Kröling, Rune Tengham.

11:30–12:10

Tal Shima (Technion—Israel Institute of Technology)

Title: Cooperative Optimal Missile Guidance Laws

Abstract: The talk is concerned with cooperative geometrical guidance of multiple missiles pursuing a single maneuvering target. An example scenario of interest is that of intercepting a high value target (such as a ballistic missile) by a team of cooperating interceptors arriving from different directions. It is customary to use in such scenarios guidance laws that have originally been developed for one-on-one engagements that can for example enforce a relative geometry between each missile and the target. In this talk optimal control based cooperative guidance laws, that enforce at intercept a relative geometry in-between the group of missiles, will be presented. The problem is posed in the linear quadratic framework and closed form analytic solutions are obtained for any team size with any linear missile dynamics. The performance of the cooperative guidance laws is investigated using a nonlinear two dimensional simulation of the missiles' lateral dynamics and relative kinematics. It will be shown that cooperatively imposing a relative intercept angle between the missiles provides substantially better results than when each missile independently enforces, using a one-on-one strategy, a preselected intercept angle that satisfies the


Abstracts | Monday, November 10, 2014relative intercept requirement. It will also be shown that the missiles' acceleration requirements are comparable to conventional guidance laws, that do not impose an

angular constraint at interception.My collaborator in this work was Dr. Vitaly Shaferman.


Abstracts | Tuesday, November 11, 20149:00–9:40

Ilana Nisky (Ben-Gurion University of the Negev)

Title: Surgical Robotics—Enhancing the Surgeon in the Loop

Abstract: Robotic manipulators are widely used in a variety of medical applications, including surgery and rehabilitation. In most cases, these devices are operated by humans (physicians) and act on humans (patients). Robotic devices are also used to study the human sensorimotor system, and their use has led to the development of rigorous computational models and theories of human motor control. However, the design and control of medical robots is rarely informed by these models and theories, thus impeding the realization of medical robots' full potential. For example, even though surgeons rely strongly on their sense of touch during open surgery, state-of-the-art teleoperated robot-assisted surgery systems do not provide them with touch information. This and other current gaps may be closed with human-centered approaches to control and design of medical robots.

In robot-assisted surgery, computational modeling of surgeons' movements combined with theories of motor coordination and learning could lead to the optimization of robot design, the development of novel teleoperation controllers, and the development of improved training curricula; all these will expand the current capabilities of robotic surgery. Towards constructing a computational model that can capture the salient characteristics of the surgeon's motor system, we studied teleoperated and freehand movements of experienced robotic surgeons and novices. We demonstrated that manipulator dynamics affect movement trajectories, and that these effects depend on expertise and on the direction of movement, suggesting that they can be modeled as a result of interplay between the dynamics of the master manipulator, the arm of the user, and neural control strategies. We also found that experienced surgeons coordinate their arm joint angles to stabilize hand movements more than novices, and that the effect of teleoperation depends on experience—experts increase teleoperated stabilization relative to freehand whereas novices decrease it.

These studies are part of a research framework that applies neuroscience to solve challenging problems in medical robotics, studies users interacting with medical robots to advance neuroscience, and employs both to improve the quality of life for patients.

9:40–10:20

Maxim Kristalny (Technion—Israel Institute of Technology)

Title: Optimal Control in Delayed Bilateral Teleoperation

Abstract: Teleoperation systems are used to expand an operator's ability to unreachable places in such areas as nuclear power, space and deep sea exploration, medical surgery, etc. Due to the very nature of teleoperation, time delays associated with communication between the local and the remote sites are ofter inevitable and can not be neglected. Handling the delays is especially problematic in the case of bilateral teleoperation, where the measurements are communicated in both directions to allow haptic feedback. In this case, both “master” and “slave” devices should be considered as dynamical systems. The need to coordinate their behavior falls into a category of challenging control problems with information constraints. In this talk, I will demonstrate how recent theoretical developments in the areas of distributed control and delayed systems may lead to a novel method for the synthesis of teleoperation controllers. The proposed approach reveals a convenient structure possessed by all stabilizing controllers and allows derivation of explicit sate-space formulae for the H2 optimal solution in spite of the distributed nature of the considered problem.

10:50–11:30

Daniel Żelazo (Technion—Israel Institute of Technology)

Title: Uncertain Consensus Networks: Robustness and its Connection to Effective Resistance

Abstract: The linear consensus protocol is one of the important canonical models in the study of multi-agent systems. This protocol has emerged as a necessary “inner-loop” for many coordination problems and is also used to analytically understand the interplay between graph theory and systems theory. The robustness of these protocols, however, has not received much attention in the literature. In this talk, we explore graph theoretic interpretations for robustness measures in uncertain consensus networks. Here, the uncertainties enter as additive perturbations to the weights used on the edges in the protocol. We show that the robust stability margins of such a setup are intimately related to properties of signed graphs and the notion of the effective resistance of a network. This framework is then extended to include non-linear sector bounded perturbations. Some illustrative examples are also provided.



Giacomo Como (Lund University)

Title: Resilient Distributed Routing in Dynamical Flow Networks

Abstract: Resilience has become a key issue in complex networked systems. Applications are of wide range and include complex infrastructure systems, such as data, transportation and power networks, as well as social, economic and financial networks. This seminar will focus on resilience properties of distributed routing in dynamical flow networks. We consider routing policies that are constrained on using only local information on the current state of the system. Disturbances are modeled as (possibly adversarial) reductions of the link flow capacities. These models allow for the possibility of cascading failures, as small local perturbations propagate through the network and local actions aimed at mitigating them can increase the vulnerability of other parts of the network. A class of maximally robust distributed policies is characterized and it is shown that, if the routing policies are allowed to use local information only from the links immediately downstream, then the resilience is in general strictly smaller than the min-cut capacity of the network, while there is no such gap if flow control is allowed using local information also from the links immediately upstream. Applications to urban traffic control will be discussed.

14:00–14:40

Anders Rantzer (Lund University)

Title: Scalable Control of Positive Systems

Abstract: Classical control theory does not scale well for large systems like traffic networks, power networks and chemical reaction networks. However, many of these applications can be handled efficiently using the concept of positive system, exploiting that the set of positive states is left invariant by the dynamics. Positive systems, and the nonlinear counterpart monotone systems, are common in many branches of science and engineering.

In this presentation, we will highlight several fundamental advantages of positive control systems: Verification and synthesis can be done with a complexity that scales linearly with the number of states and interconnections. Distributed controllers can be designed by convex optimization. Lyapunov functions and storage functions for nonlinear monotone systems can be built from scalar functions of the states, with dramatic simplifications as a result.

In spite of a rich set of existing results, several fundamental questions in control of positive systems remain open. For example, negative feedback can easily destroy positivity of the closed-loop system. On the other hand, intuition tells us that something is wrong with a traffic control system where fewer cars leads to more congestion. Hence, we need to better understand the limitations and potential of closed loop positive systems.

14:40–15:20

Karl Henrik Johansson (Royal Institute of Technology)

Title: Control and Optimization of Cooperative Driving for Road Goods Transportation

Abstract: Goods transportation is of outmost importance for economic growth and social development. About 45% of all freight transport is on roads and in Europe alone there are more than 2 million heavy-duty vehicles carrying out long-haulage services. Despite the influence transportation has on energy consumption and the environment, road goods transportation is mainly done by individual long-haulage trucks with no real-time coordination. In this talk, we will discuss how modern information and communication technology supports a future integrated goods transportation system where fleet of trucks are coordinated to travel together in vehicle platoons. From the reduced air drag, platooning trucks traveling close together can save about 10% of their fuel consumption. A system architecture utilizing vehicle-to-vehicle and vehicle-to-infrastructure communication enabling receding-horizon optimal control of individual trucks as well as optimized platoons and fleets of platoons will be presented. Experiments will illustrate how to optimize system performance while guaranteeing safety requirements. Results from a large-scale evaluation currently being performed on the highway road network in Northern Europe will also be discussed.The presentation will be based on joint work with collaborators at KTH and at the truck manufacturer Scania.



Jack Haddad (Technion—Israel Institute of Technology)

Title: Optimal Perimeter Control Synthesis for Two Urban Regions with Boundary Queue Dynamics

Abstract: Optimal perimeter control policy for multi-region system with MFD-based modeling has been presented in previous works. Notably, the derived optimal perimeter control policy might meter the number of transferring vehicles, resulting in queueing vehicles at regional boundaries. Concentrated vehicles at boundaries might affect the existence of well-defined MFDs, since congestion is assumed to be distributed homogeneously over the region implying well-defined MFDs. Previous works neglect the effect of the boundary concentrated vehicles on the traffic flow dynamics, and do not consider their effect on the permitter control policy.

With respect to modeling, a new hybrid network structure is developed, where the dynamic flow characteristics for two urban regions are modeled by MFD relationship, and boundary aggregated queue dynamics for both regions are modeled by input-output balance differential equations. Maximum lengths for boundary queues are imposed to prevent aggregated boundary queue spillbacks to the regional MFDs, such that the existence of well-defined MFDs and their dynamics are maintained.

With respect to control, the optimal control policy to maximize the total network throughput is found. Analytical solutions for the optimal perimeter control problem with constrained perimeter control inputs and constrained lengths of boundary aggregated queues are derived. The optimal synthesis for principal cases are found.

16:30–17:30

Karl Johan Åström (Lund University)

Title: Control—A Perspective

Abstract: Feedback is an ancient idea, but feedback control is a young field. Nature long ago discovered feedback since it is essential for homeostasis and life. It was the key for harnessing power in the industrial revolution and is today found everywhere around us. Its development as a field involved contributions from engineers, mathematicians, economists and physicists. It is the first systems discipline; it represented a paradigm shift because it cut across the traditional engineering disciplines of aeronautical, chemical, civil, electrical and mechanical engineering, as well as economics and operations research. The scope of control makes it the quintessential multidisciplinary field. Its complex story of evolution is fascinating, and a perspective on its growth is presented in this paper. The interplay of industry, applications, technology, theory and research is discussed.


Documents

2ND SWEDISH-ISRAELI CONTROL CONFERENCEiaac.technion.ac.il/SwIsr/SwIsr brochure.pdf · (lumped) electrical networks theory (R-L-C), by W. Cauer, R. M. Foster, O. Brune, and others