Available online at www.sciencedirect.com ScienceDirect Procedia Engineering 192 (2017 ) 427 432 TRANSCOM 2017: International scientific conference on sustainable, modern and safe transport The tunnel ventilation system in MATLAB in cooperation with TuSim Peter Kello a *, Jozef Hrbček a, Juraj Spalek a a Department of Control and Information Systems,Faculty of Electrical Engineering, University of Žilina, Univerzitná 8215/1, Žilina 010 26, Slovakia Abstract The paper deals with an influence of the ventilation system to the tunnel tube parts according to the traffic intensity. Simulations have been made by the PLC based tunnel simulator TuSim and the tools MATLAB and Simulink. The connection between MATLAB and TuSim has been made by free OPC server. The aim of this paper is to demonstrate dependence between base tunnel inputs like traffic intensity, lighting, atmospheric condition, air and car velocity and output pollutions. The obtained model is used to predict the output pollution values to the future. 2017 The The Authors. Published by Elsevier by Elsevier Ltd. This Ltd. is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of TRANSCOM 2017: International scientific conference on Peer-review sustainable, under modern responsibility and safe of transport. the scientific committee of TRANSCOM 2017: International scientific conference on sustainable, modern and safe transport Keywords: Tunnel simulation; simulation; ventilation; MATLAB; OPC server 1. Introduction Road tunnels are important part of a traffic infrastructure. They shorten the paths in the mountainous regions and in towns. Shorter travel times lead to higher economical effectiveness. The occurrence of traffic accidents in the tunnel is less common, but consequences can be more serious. A lot of technological equipment is necessary to provide the tunnel system safe in any circumstances. Simulation experiments concerning optimization of technological equipment and control algorithms cannot be realized during 24/7 operation in real tunnel, therefore Tunnel Simulator (TuSim) has been developed. Models can be used to simulate expected process behavior with a proposed control system. The * Corresponding author. E-mail address: peter.kello@fel.uniza.sk 1877-7058 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the scientific committee of TRANSCOM 2017: International scientific conference on sustainable, modern and safe transport doi:10.1016/j.proeng.2017.06.074
428 Peter Kello et al. / Procedia Engineering 192 ( 2017 ) 427 432 model requirements are a function of the usage of the model. The system is the tunnel tube. Emissions from cars are determined not only by the way they are built but also by the way they are driven in various traffic situations. Gases emitted by engine combustion contain mostly the oxides of nitrogen (NOx), carbon monoxide (CO), steam (H 2O) and particles (opacity). We are going to describe the dynamic behaviors of the system. Relation between a traffic intensity, a vehicle velocity and an air velocity is also needed to describe this system. In this paper we pointed on the concentration of CO, NOx and opacity inside the tunnel tube, because this type of pollution is the most dangerous for human organism. Using the model we can predict concentration of CO, NOx and Opacity. 2. Simulation tools 2.1. TuSim TuSIM is a PLC based system running on the B&R Automation embedded PC (PLC [7]) with the UPS unit. TuSIM hardware is displayed in Fig.1 from top to bottom: Masterview LCD switch, visualization server, UPS unit on the bottom left part of the figure and B&R industrial PC on the bottom right part of the figure. Fig. 1. TuSim hardware All-important devices of the tunnel technological subsystem equipment are simulated by the software inside the PLC [1]. Equipment of three tunnels is implemented: City tunnel, Motorway two-tube and Motorway one-tube tunnel all with a length of 1km. TuSim supports in addition to the simulation of the technological equipment also the control of the traffic sequences. Fig.2 shows the part of the visualization traffic screen with the status of the traffic sequence together with implemented devices of traffic control equipment. Each tunnel tube can operate in the following traffic sequences: tunnel tube open, left lane closed, right lane closed, speed limit 60km/h, accommodation lighting failure, tunnel tube closed. Switching from one sequence to another follows the time requirements which allow all vehicles to adapt to the new conditions. Control of the tunnel reflexes is the last important functionality implemented in the TuSim. Tunnel reflex is a reaction of the control system to a relevant event in the tunnel. The whole source code concept from the PLC software to the visualization screens is open for model enhancements, e.g. the traffic model can be easily implemented. There are many graphical screens to visualize the state of each subsystem of the technological equipment at least one for each subsystem. Handling of the screens and separate connections to the simulator are realized by the visualization server and two client PCs with HMI/SCADA CIMPLICITY software which uses the client/server architecture. The server is responsible for collection and distribution of the data from the PLC; clients allow interacting with the data distributed by the server and perform control actions. There is detailed description of the operation inside the TuSim in [2].
Peter Kello et al. / Procedia Engineering 192 ( 2017 ) 427 432 429 Fig. 2. Visualization traffic screen 2.2. MATLAB and Simulink MATLAB is a multi-paradigm numerical computing environment and Simulink is a graphical programming environment for modeling, simulating and analyzing multi-domain dynamic systems, which is integrated in MATLAB. We have used Simulink to a obtaining and a transferring data from/to the TuSim using OPC toolbox. OPC toolbox itself is a part of Simulink. 3. The realization of the connection by OPC Our aim was to create the connection between Simulink and TuSim to simulate the ventilation in the tunnel. The connection is realized by the free OPC server [3]. The OPC server is an application, which can communicate with the PLC from different manufactures. It contains the drivers for the communication between the PLC and the operating system and on the other hand it can share data for another application. In our case, the B&R automation PC simulates the PLC Siemens S7-400. The architecture of the TuSim is in Fig.3 a). HMI Server is replaced by KEPServerEX software [3] and the client is Simulink with the OPC toolbox (PC-client with Matlab, Simulink, OPC server (KEPServerEX) and VPN client), it is in Fig.3 b). It was very important to define all variables in KEPServerEX. For example the variables are: traffic flow, traffic intensity, CO, NO etc. TuSim works in its own virtual private network (VPN) Fig 3. It is necessarily connected into the VPN by the VPN client. The whole process of the linking to the KEPServerEX and defining the variables is published in [4]. Fig. 3. (a) Architecture of TuSim; (b) The Replacement of HMI server with PC-client. Through this approach it is possible to verify the existed model in the TuSim or create a new model like in our case.
430 Peter Kello et al. / Procedia Engineering 192 ( 2017 ) 427 432 4. System description and model A mathematical model is an abstract model that uses mathematical equations to describe the behavior of the real system. System identification is a process of model creating of the dynamic system from experimental data. To make the models we can use the parametric identification. Input values are traffic intensity (sensed by cameras [6], [11]), atmospheric pressure, velocity. Output values are CO (carbon monoxide) concentration, NOx (oxides of nitrogen) concentration and opacity inside the tunnel [5]. These data are measured directly by sensors inside the tunnel tube [9]. The tools MATLAB and Simulink are used not only for models creation but also for model validation. The main tasks of system identification were the choice of the model type and model order. We used a linear time discrete model because measured data have discrete character as well. When operated, many systems within nominal parameters had behavior that was close to linear. However the most realistic systems are non-linear. Therefore the linear time discrete model is an acceptable representation of the input/output behavior. System (S): A defined part of the real world. Interactions with the environment are described by input signals, output signals and disturbances. Data characterizing the existing ventilation system can be used to analyze and identify the significant parts of the system and to create its stochastic or deterministic models. Model (M): A description of a system. The model should capture the essential behavior of the system. In the case of the model OE (Output Error) we suppose that the stochastic component ξ(k) appears as a white noise additive to the output quantity [14]. Turbulence inside the tunnel (near to sensors), variety of traffic and atmospherics make the system behavior stochastic. For Multiple-input Multiple-output system we can write the matrix that describes the partial transfers between the particular output and input. The model is used to predict the pollutions inside the tunnel tube and effectively control the pollutions levels below the dangerous limits [10]. Effectively control means that pollutant concentration is kept within the allowable limit (for OP 5 km -1, NOx 0,5 ppm, CO 75 ppm or less), and at the same time electric power consumption is minimized. Another criterion is also possible, for example: number of switching the jet fans. 4.1. Output-error model Fig. 4. Particular system identification in Simulink The following equation shows the form of the output-error model [12], [13]: 1 Bz ( ) yk ( ) uk ( n) ( ) 1 k k Az ( ) (1) where: 1 1 n 1 n 1 n 1 n Az ( ) 1 az... a z az (2)
Peter Kello et al. / Procedia Engineering 192 ( 2017 ) 427 432 431 1 1 m 1 m 0 m 1 m Bz ( ) bz... b z b z (3) The variable y(k) is the output, u(k) is the input, n k is the system delay, ξ(k) is the system disturbance and k is the time step. The Fig. 5. shows the signal flow of an output-error model. ξ(k) u(k) + + y(k) Fig. 5. Output-error model structure For a given time step k, the estimation error is calculated as: e k y k y k s (4) 4.2. Simulations results The simulation shows the output values of pollutant concentrations only on the basis of the input variable, which is the traffic intensity. This feature realizes the test of the outputs difference between the model and the real system in a graphical way. The result is a graph of the output using the model and output measured in the real system tunnel Mrázovka (Fig. 6). The created model includes all influence the dynamic system such as the dimensions of the tunnel effect of weather conditions, and others in its coefficients. Fig. 6. Simulation using OE model (model parameters: n=4, m=4 and n k =6)
432 Peter Kello et al. / Procedia Engineering 192 ( 2017 ) 427 432 This algorithm uses a recursive identification method where we can set the time of model update. In this work the model was updated approximately each hour. 5. Conclusion We have successfully implemented the connection between Simulink and the PLC based tunnel simulator TuSim. The connection is realized by the OPC server. This way allows verification of existing models of our tunnel simulator TuSim. In the next section we have described the creation of the models in Simulink. Methodology that has been used for design parametric model of the urban tunnel system was presented. We need to identity the real system based on the data obtained from the real tunnel tube. The model from measured data has been created in Simulink and verified in the MATLAB environment. This part is the base for the best design of the ventilation control system. Presented results point out that the model created by identification can be used to predict the concentrations of pollutions. Acknowledgements This paper is the result of the project implementation: Centre of excellence for systems and services of intelligent transport, ITMS 26220120050 supported by the Research & Development Operational Program funded by the ERDF. "Podporujeme výskumné aktivity na Slovensku/Projekt je spolufinancovaný zo zdrojov EÚ References [1] J. Kopásek, SW for simulating of functionality of road tunnel technology equipment, user manual, ELTODO EG, 2013 [2] P. Holečko, J. Hrbček, R. Pirník, I. Miklóšik, J. Spalek, V. Šimák, Applied telematics selected chapters I, EDIS, ISBN:9788055409764, 2015. [3] KEPWARE Technologies, KEPServerEX free trial of OPC server, http://www.kepware.com [4] P. Holečko, J. Hrbček, R. Pirník, I. Miklóšik, J. Spalek, V. Šimák, Applied telematics selected chapters II, EDIS, ISBN: 9788055410043, 2015. [5] J. Hrbček., V. Šimák, Implementation of Multi-dimensional Model Predictive Control for Critical Process with Stochastic Behavior, chapter in: Advanced Model Predictive Control, p.109-124, InTech, Tao Zheng (Ed.), June 2011, ISBN 978-953-307-298-2. [6] P. Holečko, E. Bubeníková, R. Pirník, Communication systems in transport hybrid ITS interface. In: Proc. of 9th international conference ELEKTRO 2012: Žilina - Rajecké Teplice, Slovakia, May 21st-22nd. IEEE, 2012, pp. 292-298, ISBN 978-1-4673-1178-6. [7] K. Rástočný, J. Ždánsky, Specificities of safety PLC based implementation of the safety function. In: Proceedings of international conference applied electronics. AE 2012, Pilsen, pp. 229-232, ISBN 978-80-261-0038-6, ISSN 1803-7232. [8] J. Hrbček, Parametric Models of Real System and the Validation Tools, In: Proc. 8 th European Conference of Young Research and Science Workers in Transport and Communications technology. p. 93-96. Žilina: June, 22. 24. 2009. ISBN 978-80-554-0027-3. [9] A. Janota a kol., Aplikovaná telematika. EDIS 2015, ISBN: 978-80-554-1037-1. [10] Z. Tan, Y. Xia, Q. Yang, G. Zhou, Adaptive Fine Pollutant Discharge Control for Motor Vehicles Tunnels under Traffic State Transition. In: IET Intelligent Transport Systems, 2015,Vol.: 9, Issue: 8, 783-791 pp., ISSN 1751-956X [11] J. Halgaš, R. Pirník, Monitoring of parking lot traffic using a video detection. In: Acta Technica Corviniensis - Bulletin of engineering. Tom 8, fasc. 3, 2015, 17-20 pp, ISSN 2067-3809 [12] P. Noskievič, Modelování a identifikace systémů, Ostrava, MONTANEX, a. s., 1999. 276 p. ISBN 80-7225-030-2. [13] MATLAB help, Estimate Output-Error polynomial model using time or frequency domain data, http://www.mathworks.com/help/ident/ref/oe.html [14] R. Johanson, System modeling and identification, Prentice-Hall, 1993, 512p., ISBN: 0-13-482308-7.