International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 1

ISSN 2229-5518

FIS Based Speed Scheduling System of

Autonomous Railway Vehicle

Aiesha Ahmad, M.Saleem Khan, Khalil Ahmed, Nida Anwar and Umer Farooq

AbstractThis paper presents the design model of speed scheduling system of autonomous railway vehicle control using fuzzy inference system (FIS). Successful development of speed scheduling and maintaining system plays crucial role to make autonomous railway vehicle control system more effective under constraint of uncertain conditions. This research work emphasis to develop the speed scheduling system with capability to adjust in uncertain conditions magnificently by improving performance, stability, controllability and safety of railway vehicles and ultimately reduce the risk to meet the needs of modern trend of autonomous control system. The proposed design model is capable to successfully cope with hard conditions; junction track information (JTI), crossing gate information (CG) and track clearance (TCL), and flexible conditions; vehicle tilting (VT), track conditions (TC) and environment monitoring (EM) using fuzzy inference system with better and quicker response with human knowledge incorporation. This system will be helpful to successfully maintain the speed of railway vehicles with envi- ronment monitoring, time scheduling and minimizes the risk of overturning.

Index TermsAutonomous Railway Vehicle, Control System, Environment Monitoring, Fuzzy Inference System, Speed Scheduling, Time



—————————— • ——————————
DVANCEMENT in railway vehicles technology has been increased from last few years to facilitate pas- sengers by improving performance with speed, time scheduling, traffic control and passenger management [1]. The importance of control, management and monitoring for railway vehicles progress gradually under modern solution of embedded systems, software based computer aided control systems, sensors and data communication technologies. The design and development of agent base autonomous railway vehicle control system is considered important for flexible and well established network to enable collaboration between centralized and distributed systems of railway tracks [2]. In these autonomous rail- way control systems, agents are defined as condition monitoring units with capability to collect information independently and control system autonomously accord- ing to their design control [3]. The monitoring and control of railway vehicles are mainly focused on speed man- agement and scheduling, traffic control and time schedul- ing. Speed scheduling plays vital role for successful de- velopment of autonomous railway control system by fo-


Ms. Aiesha Ahmad is with Computer Science department as research fellow at NCBA&E Lahore, Pakistan (e-mail:

Dr. M. Saleem Khan is with the Computer Science Department as Direc- tor in GC University Lahore, Pakistan (e-mail:

Dr. Khalil Ahmed is with the School of Computer Science at NCBA&E Lahore, Pakistan; He is an expert academician and passionately engaged in research. (e-mail:

Ms. Nida Anwar is with Computer Science department as research fel- low at NCBA&E Lahore, Pakistan. She is currently working as faculty member in VU, Lahore, Pakistan (e-mail: nidaanw

Mr. Umer Farooq is with Computer Science department as research fellow and faculty member at NCBA&E Lahore, Pakistan. (e-mail:

cusing on hard conditions; Junction track information (JTI), crossing gate information (CG) and track clearance (TCL), and flexible conditions; vehicle (VT) tilting, track condition (TC) and environment monitoring (EM). To maintain information about track condition with support of acceptable ride quality and tilting of trains around curved tracks with speed adjustment have been done through some sensors, mathematical modular technique and kalman filtering for data estimation and tracking in existing systems. Mathematical modular techniques sometimes unable to meet the need of real time environ- ment due to lack of flexibility while kalman filtering tech- nique also can’t perform well in presence of noise in ini- tial stages with modeling of system [4]. The complexity and dynamic nature of autonomous railway control sys- tem is needed some sophisticated method with domain knowledge representation and save time with quick re- sponse to handle uncertain situations successfully during running on track. Fuzzy inference system has capability to perform uncertain reasoning under incorporation of human knowledge in real time environment with better and quicker response. The proposed design model will be capable to adjust speed under uncertain conditions with high safety, performance and time management.

The arrangement of this research paper is as follows: section 2 consists of brief overview of fuzzy logic and fuzzy inference system that helps to understand the im- portance of fuzzy base systems. Structure of proposed speed scheduling system is discussed in section 3. Design algorithm is presented in section 4 and results and dis- cussion is described in section 5 on the basis of design algorithm. Section 6 presents conclusion and future work on the basis of design algorithm.

IJSER © 2011

International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 2

ISSN 2229-5518



Imprecise modes of reasoning are successfully demon- strated by fuzzy logic with appropriate answer that con- sidered fundamental aspect of human rational decision making under ambiguous and uncertain environment based on incomplete, inexact or bit reliable knowledge [5], [6], [7], [8]. Fuzzy logic simulates human reasoning through mathematical theory of fuzzy sets with power of high precision while mapping of input values to an out- put values is done using process of fuzzy inference which drives to the final decision. Therefore fuzzy inference sys- tem (FIS) is considered as one of main computing frame- work in artificial intelligent (AI) that integrate human knowledge with idea of fuzzy IF-THEN rules to deter- mine imprecise and uncertain reasoning in real time envi- ronment. FIS provides better, quicker and more appropri- ate solution as compared to traditional approaches be- cause these approaches worked with crisp set of distinct and precise boundaries while in fuzzy set transition from non-membership to membership is gradual.



The fuzzy inference based proposed speed scheduling system consists of preloaded information of track in form of root chart and design to cope with uncertain conditions successfully with minimum response time. Railway track system receives two main conditions (hard and flexible) from environment and uses sensors to differentiate be- tween these conditions. In case of flexible conditions (FC): VT, EM and TC observed through sensors and given to FIS. These sensors are capable to monitor EM, TC and VT individually by subdividing into 0 to 5 volt in which 0 volt represents the absence of these flexible conditions with no need of speed adjustment while from 1-5 volt shows gradual increase in these conditions with respec- tive decrease in speed, like 5 volt represents worst condi- tion against one or more flexible conditions (VT, TC or EM) with slow speed. Hard conditions (HC) like JTI, TCL and CG are sensed through sensors at particular distance and ultimately stop the train whatever adjusted speed may be after existence of flexible conditions. Fig.1 is representing the basic structure of FIS based speed sche- duling system in presence of both FC and HC using fuzzy logic.
The proposed system will be capable to reduce speed or stop by increasing time according to requirement in uncertain environment. Then it will compare the increase time with root chart estimated time to reach next junction and increase speed to its maximum possible limit to over- come the delay after handling uncertain situation.

Fig.1. Block Diagram of Speed Control System using FIS.

The design procedure can be explained with the help of mathematical equations.
In the start, speed is
S = V * t
After uncertain conditions, the speed reduced as
S = (V + LV) t
Change in velocity calculated as
S / t – V = LV
Then this change in speed will compare with root chat estimated speed to overcome the delay to reach the next junction according to root chart calculated time. An over- view of speed scheduling system with fuzzy control and speed adjustment is given in Fig. 2.

Fig.2. Block Diagram of Speed Scheduling System.

Fuzzy control system is used to adjust speed quickly and precisely in the presence of any flexible condition or delay due to any hard condition after comparing with root chart. The fuzzy control system for proposed speed scheduling system consists of fuzzifier, inference kernel with knowledge base including database, rule base and output membership functions, and defuzzifier block as shown in Fig.3. The crisp values of input variables VT, TC and EM are reached to fuzzifier after passing through sensors to identify the types of these input variables [7], [9], [10]. In fuzzifier, comparison of input crisp values up to certain levels is done by generating linguistic values (Low, Medium, Average, High and Very High) against each input variable. These linguistic values are passed to inference kernel connected with knowledge base which further categorize into database, rule base and output membership functions.

IJSER © 2011

International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 3

ISSN 2229-5518

In knowledge base, key feature of database is to ma- nipulate fuzzy data and provide essential definitions to describe the linguistic control rules which help the rule base to define the control goals and control policy of par- ticular system such as speed scheduling of railway ve- hicle in this scenario while output membership functions define the strength of output variables with formulation to adjust speed. After receiving feedback from knowledge base, the next step of inference kernel is to simulate the human decision with fuzzy logic rules to make the control decision in term of adjusted speed, the final outcome. In the next step, defuzzifier maps fuzzy output variable (Slow, Average, Fast and Very Fast) to a crisp value which finally comes to railway track system after passing through actuator.

Fig.3. Design Model of FIS Based Speed Scheduling System.


The proposed model is designed for three input variables (VT, EM and TC) which are derived from flexible condi- tions. The membership functions for these three input variables are shown in Table1.



The five membership functions mf1 [1], mf1 [2], mf1 [3], mf1 [4] and mf1 [5] are used to represent the different ranges of input fuzzy variable “Vehicle Tilting” as shown
in Fig. 4. The plot of VT consists of four regions. The other two variables, EM and TC are design on the same pattern for simplification and better understanding.

Fig. 4. Plot of Membership Functions for Input Fuzzy Variable- Ve- hicle Tilting.

The output variable speed consists of seven membership functions. The detail about each membership functions with scale and singleton values are shown in Table 2. The plot of seven membership functions of speed with maximum speed limit 120km/h is shown in Fig. 5.



Fig. 5. Plot of Membership Function for output Fuzzy Variable- Speed.

4.1 Fuzzification

There are three fuzzy input variables in proposed design model of speed scheduling and each variable is divided into four regions while f1 and f2 are linguistic values of fuzzy variable “Vehicle Tilting”, f3 and f4 for “Environ- ment Monitoring” and f5 and f7 for “Track Condition”.

IJSER © 2011

International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 4

ISSN 2229-5518

The linguistic values described as the mapping values of these three fuzzy input variables: VT, EM and TC with their membership functions categorize into four regions as shown in Table 3.



Input variables, VT, EM and TC are inversely propor- tional to speed while each variable has independent effect on speed, even with minor change in any one variable. The number of rules for complete simulation of design model is 64 and rule base will maintain the record of these rules. In this case, 8 rules are used for the values of specific region of input variables like VT=10, EM=15 and TC=25 are taken for region 1 with membership functions and corresponding mapping values mf1 [1], mf1 [2], mf2 [1], mf2 [2], mf3 [1], mf3 [2]. The Fig. 6 is representing the fuzzi- fication process of these input variables crisp values to lin- guistic variables.

Fig. 6 Fuzzifier Representing, 3- Inputs- Crisp Values with 6- Out- puts- Linguistic Variables.

4.2 Fuzzy Inference Engine

The inference engine consists of eight AND operators which select minimum value from linguistic values of three input variables. The working technique of these AND operators are different form logical ANDs. This inference engine accepts six inputs from fuzzifier and ap-

plies min-max composition to obtain the value to adjust speed. The min-AND operation is used from min-max tech- nique to get minimum value from input variables TC, EM and TC as shown in Fig. 7 which gives the overview of fuzzy inference engine working strategy.

Fig. 7. Block Diagram of Fuzzy Inference Engine.

Rule1 = f1 ^ f3 ^ f5 = mf1 [1] ^ mf2 [1] ^ mf3 [1]

= 0.33 ^ 0.5 ^ 0.83 = 0.3

Rule2 = f1 ^ f4 ^ f6 = mf1 [1] ^ mf2 [2] ^ mf3 [2]

= 0.33 ^ 0.5 ^ 0.16 = 0.16

Rule3 = f1 ^ f3 ^ f6 = mf1 [1] ^ mf2 [1] ^ mf3 [2]

= 0.33 ^ 0.5 ^ 0.83 = 0.33

Rule4 = f1 ^ f4 ^ f5 = mf1 [1] ^ mf2 [2] ^ mf3 [5]

= 0.66 ^ 0.5 ^ 0.83 = 0.5

Rule5 = f2 ^ f3 ^ f5 = mf1 [2] ^ mf2 [1] ^ mf3 [1]

= 0.66 ^ 0.5 ^ 0.83 = 0.5

Rule6 = f2 ^ f4 ^ f6 = mf1 [2] ^ mf2 [2] ^ mf3 [2]

= 0.66 ^ 0.5 ^ 0.16 = 0.16

Rule7 = f2 ^ f3 ^ f6 = mf1 [2] ^ mf2 [1] ^ mf3 [2]

= 0.66 ^ o.5 ^ 0.16 = 0.16

Rule8 = f2 ^ f4 ^ f5 = mf1 [2] ^ mf2 [2] ^ mf3 [1]

= 0.66 ^ o.5 ^ 83 = 0.5

The ^ operator between membership function values is used for min-AND process to get minimum value of the function.

4.3 Rule Selector

The rule selector of proposed model receives three crisp values of VT, EM and TC, and provides singleton values of output function with specific rules of the design model re- quirement. In this case, 8 rules are required to find the re- quired values S8, S9…, S15 according to division of regions for soft conditions while hard conditions are absent.

The rules are listed in Table 4 with existence of both hard and flexible conditions for quick overview of rules for speed control system.

4.3 Defuzzifier

In defuzzification process, crisp values for final estimated speed are obtained after estimating its inputs regarding

IJSER © 2011

International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 5

ISSN 2229-5518

TC, EM and VT from the rule base.



There are 16 inputs are given to the defuzzifier, eight values from eight rules (Rule1-Rule8) and eight values from rule selector (S8-S15). The center of average method (C.O.A) is used by each defuzzifier to estimates the crisp value with mathematical expression .L Si * Ri / .L Ri, where i = 1 to 8 in the given scenario as shown in Fig. 8.

Fig. 8. Block Diagram of Defuzzifier.

The mathematical expression .L Si * Ri / .L Ri has used for crisp values of output variables which are according to MATLAB simulation results as shown in Fig. 9.
The values of input variables, VT, TC and EM are taken as the same for MATLAB simulation as for mathematical calculation that shows correct results with quick re- sponse.

Fig.9. MATLAB- Rule Viewer and Simulation Results for Speed

Scheduling System of Railway Vehicles.


The design model of proposed speed scheduling system of autonomous railway vehicle has shown significant im- provement in performance regarding safety and time to meet uncertain conditions with minimum delay using fuzzy inference system with MATLAB simulation as compared to traditional approaches. Soft conditions like VT, EM and TC are inversely proportional with respect to speed and have shown substantial effect on it in case of any change in input variables individually and combined as well. The effect of these inputs on speed has shown in Fig.10 which is according to rule base of design algo- rithm.
Fig. 10 (a) has shown that at 0 of both VT and EM, the speed is fast, more than 100 km/h which gradually de- crease with increase in values of EM and VT.
Fig. 10 (b) has shown the same effect with input va- riables VT and TC which is according to rule base by as- signing specific values to input variables to specific re- gion and prove that these three variables have same effect on speed.

Fig. 10 (a) Plot between Environment Monitoring and Vehicle Tilting.

IJSER © 2011

International Journal of Scientific & Engineering Research Volume 2, Issue 6, June-2011 6

ISSN 2229-5518

Fig. 10 (b). Plot between Vehicle Tilting and Track Condition.


FIS based speed scheduling system of autonomous rail- way vehicles has shown remarkable improvement to compete the demands of new trends with safety and min- imum time delay. Now day’s railway control system re- quires some sophisticated method to handle real time problems without compromising on schedule time and security. MATLAB results have shown that proposed model of railway speed scheduling system will be capable to handle uncertain conditions successfully with better and quicker response as compared to existence methods with time management and safety by reducing risk. In future, fuzzy inference system base railway control sys- tem will be more secure with performance enhancement in real time environment. State of the art Microelectronics technology can be used to develop FPGAs based control chips for this autonomous railway control system.


This research work was carried out in the laboratories of NCBA&E and GC University, Lahore, Pakistan. We must acknowledge the support of research group fellows and laboratories personals for their supporting and encourag- ing behavior.


[1] M. Saleem Khan and, Khalid Benkrid, “Design Model of Multi Agents Based Autonomous Railway Vehicles Control Sys- tem”, in proc. Of int. Multiconference of Engineering and com- puter Scientists, Hong Kong, March 18-20, 2009, vol.2.

[2] R. Harrison, M. Lee, “Reconfigurable Modular Automation Systems for Automotive Power-train Manufacturer”, int. Flex Manuf Syst, Springer Science 1 March 2007.

[3] M. Saleem Khan and, Khalid Benkrid, “Development Tech- niques of Multi-Agents Based Autonomous Railway Vehicle Control System”, World Academy of Science Engineering and Technology, pp. 558-568, 2010.

[4] M. Malleswaran, J. Mary Anita, S. N. Sabreen and V. Vaidehi

“Real-time INS/GPS Data Fusion Using Hybrid Adaptive Net- work Based Fuzzy Inference”, 11th Int. Conf. Control, Au- tomation, Robotics and Vision Singapore, pp. 1536-1540, De- cember 2010.

[5] Lofti A. Zadeh, “Fuzzy Logic”, pp. 83-92, IEEE, April 1988.

[6] S.H. Lee, R.J. Howlett and S.D. Walters, “Fuzzy and Neuro- fuzzy Techniques for Modeling and Control”, pp.1206-1215,

Springer-Verlag Berlin Heidelberg 2006.

[7] Tang Xianzhi and Wang Qingnian, “Driving Intention Intelli- gent Identification Method for Hybrid Vehicles Based on Fuzzy Logic Inference”, Int. Symposium on Computational Intelli- gence and Design, pp. 16-19, IEEE, 2010.

[8] Gary R. George P. E. and Frank Cardullo, “Application of Neu-

ro-Fuzzy Systems to Behavioral Representation in Computer

Generated Forces”.

[9] Morelli, Jordan E., Akira Hirose and Hugh C. Wood “Fuzzy- Logi-Based Plasma Position Controller for STOR-M” IEEE Transaction on Control Systems Technology, Vol. 13, No 2, March 2005.

[10] M. Saleem Khan, “Fuzzy Time Control Modeling of Discrete Event Systems”, Proc. Of World Congress on Engineering and Computer Science 2008.

IJSER © 2011