1,721,048 research outputs found
Model predictive control for ramp metering of motorway traffic: A case study
No full textA real-life motorway in Belgium is studied and a comparison is made between a simulation of a morning rush hour situation without control and a simulation of a morning rush hour situation with ramp metering implemented. Two types of controllers are compared: a traditional ALINEA based controller and a model predictive control based ramp metering controller. In order to evaluate the controllers in a realistic framework, the simulations presented in this paper are based on real-life traffic measurements, and constraints on the maximal allowed queue lengths at the on-ramps are imposed. The presented simulations are indicative for the reduction in the total time spent (on the studied motorway and on the on-ramps) that can be achieved by ramp metering during a typical morning rush hou
A traffic responsive control framework for signalized junctions based on hybrid traffic flow representation
No full textThe paper proposes a traffic responsive control framework based on a Model Predictive Control (MPC) approach. The framework focuses on a centralized method, which can simultaneously compute the network decision variables (i.e., the green timings at each junction and the offset of the traffic light plans of the network). Furthermore, the framework is based on a hybrid traffic flow model operating as a prediction model and plant model in the control procedure. The hybrid traffic flow model combines two sub-models: an aggregate model (i.e., the Cell Transmission Model; CTM) and a disaggregate model (i.e., the Cellular Automata model; CA), using a transition cell to connect them. The whole framework is tested on a signalized arterial, performing several analyses to calibrate the MPC strategy and evaluate the traffic control approach using fixed and adaptive control strategies. All analyses are made in terms of total time spent, network total delay, queue lengths and degree of saturation
Modeling and Control of Switching Max-Plus-Linear Systems: Rescheduling of railway traffic and changing gaits in legged locomotion
No full textThe operation of many systems can be described by the timing of events. When the system behavior can be described by equations that are "linear'' in the max-plus algebra, which has maximization and addition as its basic operations, the system is called a max-plus-linear system. In many of these systems the order of the events may need to be changed due to changes in the conditions, or the requirements. Such systems that can change the order of events are called switching max-plus-linear (SMPL) systems. In this thesis we consider two application of SMPL systems. The first application of SMPL systems models the railway traffic networks and is used for on-line rescheduling of railway traffic in the case of delays. In this thesis a macroscopic model for the railway traffic network is presented that can model the effects on the railway traffic of several control actions. For every set of control actions the new event order and times are determined. In order to solve the on-line rescheduling problem for a railway traffic network a global model predictive control (MPC) approach and four distributed model predictive control (DMPC) approaches are proposed. The second type of SMPL system models legged locomotion for different gaits. In this thesis the steady state cyclic behavior of the max-plus-linear systems describing the gaits, and the transition to the steady state cyclic behavior, are analyzed. It is shown that the steady state behavior can be uniquely defined for all gaits. With the steady state behavior uniquely defined for all gaits we were able to determine optimal gait switches.Delft Center for Systems and ControlMechanical, Maritime and Materials Engineerin
Approximation Methods in Stochastic Max-Plus Systems
No full textStochastic max-plus systems belong to a special class of discrete-event systems. This class consists of systems with synchronization but no choice and the models of such systems are defined using the operators maximization and addition. Stochastic max-plus systems can be further extended to stochastic switching max-plus systems and stochastic min-max-plus-scaling systems. In the identification and control problem of all these systems, the objective function appearing in the optimization problem can be written as the expected value of the maximum of several affine expressions. The focus of this thesis is on finding an efficient method to compute this expected value since the currently available methods are both too complex and too time-consuming. To address this issue, this thesis proposes an approximation method based on the higher-order moments of a random variable. By considering the relationship between the infinity-norm and the p-norm of vectors, we obtain an upper bound for the expected value of the maximum of several affine expressions. This approximation method can be applied to any distribution that has finite moments and in the case that these moments have a closed form (such as for a uniform distribution, normal distribution, beta distribution, or gamma distribution), the approximation method results in an analytic expression. For all the above-mentioned systems, we have compared the performance of the proposed approximation method with other available methods, such as analytic and numerical integration, and Monte Carlo simulation. In nearly all cases, the computation time of the proposed approximation method is at least two orders of magnitude smaller than that of other methods, while still resulting in a comparable control performance.Delft Center for Systems and ControlMechanical, Maritime and Materials Engineerin
ARSOn: A Robotic Search Optimization
No full textRobotic search is a very active field of research, and especially search with multiple robots is of high interest. A swarm of robots equipped with sensors could be used in a variety of useful settings such as border patrol, monitoring water quality of the ocean with underwater robots, or – in case of the FireSwarm project – for finding dune fires with Unmanned Aerial Vehicles (UAVs). This thesis was assigned by the research company Almende B.V. and is related to the FireSwarm project. The main idea behind projects like FireSwarm is that a large group of cheap UAVs with cheap less reliable sensors can search more effectively than one expensive UAV with more reliable sensors. Therefore, the main focus in this thesis is on the development of efficient robotic search strategies for different sensor representations. The robotic search problem that is defined in this thesis is the problem of finding an important point (a fire) in a predefined search area as fast as possible with autonomous UAVs. Herein, a clear distinction is made between search with perfect (deterministic) sensors and search with more realistic and imperfectly (stochastically) modeled sensors. The modeling of the fire sensors is very general in order to be able to represent any kind of (fire) sensor. This is achieved by, instead of qualitatively modeling the effect of each factor on the fire sensor, modeling the sensor as a simple stochastic process. Two main conclusions can be drawn from the tests performed in this thesis. First, the optimal strategy for search with a single UAV, with perfect deterministic sensors, is shown to be a predefined sweeping path. Furthermore, the methods proposed in this thesis for predefining a search path for multiple UAVs with deterministic sensors resulted in near- optimal performance. Secondly, it is concluded that for increasingly stochastic sensors, the predefined sweeping path is easily outperformed by the simple greedy strategy proposed in this thesis. So, because of the general representation of the search problem, the results found in this thesis cannot only contribute to the development of fast efficient algorithms for the FireSwarm project but also to other real-world (robotic) search problems.BMDBioMechanical EngineeringMechanical, Maritime and Materials Engineerin
Cooperative adaptive cruise control: Using information from multiple predecessors in combination with MPC
No full textCooperative adaptive cruise control (CACC) makes the vehicle follow its predecessor at a close but safe distance, and uses information received from other vehicles to accomplish this task. In literature and in practice, the control method mostly applied for CACC is proportional integral derivative (PID) control, possibly with some refinement for gear shifting or comfort. The control method called model predictive control (MPC) can also be used for CACC, and from literature it appears to be more promising than PID, because of its ability to anticipate future situations and to implement constraints directly into the control algorithm. MPC applies the first input of a control input sequence that optimises a performance index calculated from predicted system behaviour, based on a prediction model, subject to operational constraints, in a receding horizon approach. Furthermore, literature has shown that with PID the use of state information from the second predecessor or the platoon leader, in addition to the direct predecessor’s states, can improve the CACC performance. Therefore, in this thesis the approach of using such additional communicated information from either the second predecessor or the platoon leader is combined with the use of MPC as control method. The goal is to investigate whether any of these two configurations give an increase in performance compared with similar configurations with PID as control method, and compared with a more basic configuration that uses just the direct predecessor’s state information with either MPC or PID. Also, the possibly added value of using communicated predicted states, in addition to current states, with MPC is investigated. The CACC controllers are designed to control the throttle, the brakes, and the gears, subject to operational constraints on acceleration, velocity, and vehicle-to-vehicle distance. The PID-based CACC controller contains a proportional feedback of the errors in velocity, position, and acceleration, combined with an automatic transmission scheme, and the control input is restricted at time instants at which a constraint is (almost) violated. The MPC-based CACC controller at each time step minimises the expected errors in position and velocity and the corresponding input variation. The MPC prediction model is obtained by approximating a nonlinear vehicle model by a piecewise affine (PWA) model, and converting the MPC optimisation problem into a mixed integer linear programming (MILP) problem. In this project, tuning is done by applying simulated annealing for a scenario involving four CACC-controlled vehicles following a platoon leader. Then, the tuned controllers are implemented in a validation scenario comprising a larger platoon undergoing a longer simulation. The results from simulating this validation scenario show that the PID-based CACC controller has a low responsiveness, compared with MPC, and lets the first two vehicles crash. With MPC several peaks and oscillations in throttle/brake input and acceleration occur, and it is expected that with the MPC-based CACC controllers as designed and tuned here, string stability will not always be achieved for increasing platoon lengths. It is expected that properly retuning will result in better performing controllers. However, due to limited time, this retuning could not be performed within the scope of this project, and is therefore left as a recommendation. Therefore, only preliminary conclusions can be formulated, which are that MPC should be preferred over PID as a control method for CACC, because it is safer. Moreover, with MPC it should be preferred to, in addition to the current states of the direct predecessor, at least use the current states of the second predecessor and/or the predicted future states from the direct predecessor, in order to achieve better string stability.Delft Center for Systems and ControlMechanical, Maritime and Materials Engineerin
Batch Scheduling of Multi-Product Pipeline Networks
No full textIn oil supply chains, crude oil needs to be transported from oil fields to refineries, and refined products need to be transported from refineries to regional depots. On land, pipelines are the preferred mode for long-distance oil transportation, because they are safe, efficient, silent, and cheap compared to other modes of transport. Pipelines are often part of large networks, in which they connect multiple supply and demand locations. Multi-product pipelines transport batches of different products, such as gasoline, diesel, and jet fuel. Pipeline networks should be operated such that temporal and spatial differences between supply and demand are balanced, operational limitations are satisfied, and costs are minimized. This is a rather complicated task, due to the size and complexity of pipeline networks, limited capacities of tanks and pipelines, and the existence of transportation times of several days. Because batches are pushed through pipelines, transportation times of current batches depend on injections of future batches, which is a distinctive feature of the pipeline scheduling problem. The minimization of operational cost is mainly related to transmix volumes, i.e. contaminated volumes that emerge between consecutive batches, and pumping energy. In this thesis, we propose a novel pipeline scheduling method for solving the pipeline scheduling problem. It consists of a planning and a scheduling phase that are coupled in a hierarchical decomposition scheme. In the planning phase, global day-to-day transportation volumes are determined for each pipeline. In the scheduling phase, we use the planning output to generate complete schedules. Both phases contain a discrete-time Mixed Integer Linear Programming (MILP) problem. The MILP planning problem is solved with truncated branch and bound. The MILP scheduling problem is further decomposed using a rolling-horizon approach; the resulting subproblems are solved with branch and bound. The pipeline scheduling method has been successfully tested on two case studies involving up to 4 products, 8 pipelines, 8 tank farms, 2 supply locations, and 5 demand locations. The proposed method is flexible in terms of network configurations, intermediate supply and demand requirements, and cost structures. Complete schedules for 30-day horizons are obtained within 3 to 4 minutes of computation time. With respect to current industry practice, the novel pipeline scheduling method can greatly reduce the time required to generate schedules. Compared to current spreadsheet approaches, the proposed method is generic and less error-prone. Moreover, the obtained schedules are significantly better in terms of transmix and pumping costs.Mechanical, Maritime and Materials EngineeringDelft Center for Systems and Control (DCSC
Controller synthesis using interval methods
No full textThis thesis investigates whether interval methods can be employed in the construction of a novel controller synthesis algorithm based on backward induction. Interval methods are methods employing interval arithmetic, which is an arithmetic defined on real-valued intervals rather than on real-valued numbers. In the controller synthesis algorithm presented in this thesis interval methods are used to determine pre-images, represent approximations of closed sets, implement operations on these sets, and solve non-linear constrained optimisation problems without the need for derivatives. While interval methods only impose modest requirements, i.e., they require that interval extensions of the difference equation describing (or approximating) the plant dynamics, cost function, and inequality constraints can be constructed, they do however suffer from the curse of dimensionality. In the presented synthesis algorithm the curse of dimensionality limits practical use to systems for which the number of states and control inputs are relatively low. The thesis can be divided into four parts: - The first part of this thesis (Chapters 2, 3 and 4) introduces interval arithmetic, a number of interval methods, and set computation. - In the second part of the thesis (Chapter 5) the controller synthesis algorithm is presented and implemented using the concepts presented in the first part of the thesis. - In the third part of the thesis (Chapter 6) the implemented synthesis algorithm is successfully used to generate, and test the viability of, controllers for two benchmark problems. - The fourth part (Chapter 7) concludes the thesis, gives recommendations for improving the synthesis algorithm and suggests a number of topics worth considering for future research. In conclusion, this thesis shows that interval methods can be used to construct a controller synthesis algorithm for non-trivial control problems.Systems & ControlDelft Center for Systems and ControlMechanical, Maritime and Materials Engineerin
Switched LQR control: Design of a general framework
No full textThis thesis studies the Switched Linear Quadratic Regulator (SLQR) problem, over a hybrid (continuous and discrete) dynamical model known as "switched system". The problem is defined as computing the optimal continuous and discrete switching control to minimise a quadratic cost function that weights the states and the continuous controls. The original SLQR problem does not handle constraints on states, continuous or discrete controls, and there is no probabilistic behaviour. This thesis focuses on the discrete dynamics in a SLQR problem. The first part of the thesis describes the SLQR problem with discrete constraints, whereas the second part is dedicated to probabilistic switching behaviour. The problem with discrete constraints is described as finding the optimal hybrid switching policy that minimises a quadratic cost function, weighting states and continuous controls, without violating the discrete constraints. The problem with probabilistic switches is defined as finding the optimal hybrid switching policy that minimises an expected value of a quadratic cost function, weighting states and continuous controls. For the SLQR problem with discrete constraints a general relaxation framework is developed to simplify the representations of the value functions and the corresponding control strategies. It is shown that the closed loop performance of the obtained solution with the relaxation framework can be made arbitrarily close to the optimal solution. For the SLQR problem with probabilistic switches it is shown that a relaxation framework can only be developed when there are no discrete constraints involved. Finally, the thesis concludes with a few case studies to illustrate how the optimal hybrid control sequence is computed.Delft Centre for Systems and ControlMechanical, Maritime and Materials Engineerin
- …
