Design Car Braking System Using Mamdani Fuzzy

2017 4th International Conference on Electric Vehicular Technology (ICEVT) October 2-5, 2017. Bali, Indonesia

[ABS-57]

Design Car Braking System Using Mamdani Fuzzy Logic Control Illa Rizianiza

Alfian Djafar

Department of Mechanical Engineering Kalimantan Institute of Technology, Balikpapan, Indonesia National Center for Sustainable TransportationTechnology, Indonesia [email protected]

Department of Mechanical Engineering Kalimantan Institute of Technology, Balikpapan, Indonesia National Center for Sustainable TransportationTechnology, Indonesia [email protected] uncertainties and nonlinierities system such as braking system [3]. Mamdani is widely accepted for capturing expert knowledge [4]. Due to the interpretable and intuitive nature of the rule base, Mamdani-type is widely used in particular for decision support application

Abstract— Car braking system is the most important safety system that it has. This paper provides an automatic car braking system using Mamdani Fuzzy Logic Control. This car braking system consist of two inputs and one output. Inputs are car position and car velocity. Car position represents the distance of the car from the obstacle detected and velocity represents the velocity of the car towards the obstacle. This output is brake that represents the car force to stop the car. The inputs use five membership funtions and the output use three membership functions. The car braking system designed uses twenty five rules. Braking performance will be observed in term of distance and velocity to prevent collision.

II. DESIGN CAR BRAKING SYSTEM A. Car Braking System In this paper, the car braking system uses the fuzzy logic to to apply the brake when the position of the obstacle is near and the velocity of the vehicle is high. The car velocity is range 0–100 km/hr. It defined based on Indonesia Government Regulations Number 43 Year 1993 that the the vehicle velocity maximum is 100 km/hr (37.7 m/s). In this car braking system, there are some parameter has been negleted. There are engine dynamic, skidding and friction [2].

Keywords— braking system; mamdani; fuzzy logic control.

I. INTRODUCTION Currently the number of car users is increasing. It causes the opportunity for the occurrence of accidents also increased. One of the causes an accidents is collision between vehicles. The reasons is the negligence of the driver when driving a car. The driver cannot control the speed of the car and the inaccuracy braking. Moreover if the car driving at high speed, it will be difficult to avoid obstacles that exist in front of it. Close to 68% people die in Indonesia from 2010 until 2016 that caused by the collision [1]. Car braking system is the most important. Generally car braking system is operated manually as the driver pushes the brake pedal. If the brake fails, the result can be failed too. In recent years, there are some research to develop car braking system. Braking developments have led to significantly greater driving safety [2]. There are many studies about car braking system depend on mathematic modeling, but the fact behaviors of drivers are mostly depend on experience not exact mathematic model. The vehicle is nonlinear system so it too difficult to find a mathematic model. Therefore, fuzzy logic has been used to develop automation control since fuzzy emulates the performance of a skilled human operator in the linguistic tulles that do not need use a mathematic model. Fuzzy system is higly robust and effective to handling the

Parameter value of the car braking system: Mass of the car : 2000 kg Car position to start brake : -30 m Initial car velocity : 10 km/s (2.7 m/s) Car braking force : 9263 N The determining of car position to start brake is based on Government Regulation Number 55 Year 2012 about Vehicle. It is a safe driving distance. According to Newton’s Second Law of Motion that the force causes acceleration, so in this paper uses the assumption that the automatic braking system limits the magnitude of the brake force to 9263 N. B. Mamdani Fuzzy Control The car breaking system is designed by Mamdani Fuzzy Logic Control. Mamdani is one of the type of fuzzy inference method. In the mamdani inference, the consequent of If-Then rule is defined by fuzzy set. The output fuzzy set of each rule

978-1-5386-3698-5/17/$31.00 ©2017 IEEE 129


[ABS-57]

will be reshaped by a matching number, and defuzzification is required after aggregating all of these reshaped fuzzy sets.

• Fuzzifying the inputs using the input membership functions • Combining the fuzzified inputs according to the fuzzy rules to establish a rule strength • Finding the consequence of the rule by combining the rule strength and the output membership function (implication). In this paper used Min implication. • Combining the consequences to get an output distribution (aggregation). In this paper used Max Aggregation. • Defuzzifying the output distribution. In this paper use Center of Area (COA) defuzzifying method.

Fig 1 Block Diagram for Car Braking System Control Car braking system is designed using closed loop control is shown at Figure 1. Input Fuzzy Logic Control are car position and car velocity and the output is brake force. Car position is the distance of the car to the obstacle detected and the car velocity is velocity of the car towards to the obstacle. The output is car brake force. Brake force is the car force to braking to stop the car. input Fuzzifier

Rule Base Fig 3 Fuzzy Logic Car Braking System Membership function is a curve that defines how each point in the input space is mapped to a degree of membership between 0 and 1. The membership function of variables are shown at Figure 4, 5, and 6. The maximum and minimum range inputs for fuzzy logic is defined based on the Government Regulation about vehicle and driving. The output of the fuzzy logic is a signal control will be the input for the car as a brake force. Where else, output variables are 3 that are Max, Zero and Min. In order to improve fuzzification speed, Gaussian function is chosen as the membership function [5].

Defuzzifier

output Fig 2 Membership Function of Car Position Variable Fuzzy inference system is a computing frameworks based on the concepts of fuzzy set theory. It can applied in many fields such as control, decision support, system identification, prediction, etc. It is mainly due to the closeness to human perception and reasoning, as well as their intuitive handling and simplicity, which are important factors for acceptance and usability of the systems [5]. Figure 2 showing that there are three main modules are of particular interest: fuzzifier, rule base and defuzzifier. Fuzzifier converts the crisp input to a linguistic variable using the membership functions stored in the fuzzy knowledge base. Defuzzifier Converts the fuzzy output of the inference engine to crisp using membership functions analogous to the ones used by the fuzzifier.

Membership function of car velocity are : Very Near (VN) Near (N) Medium (M) Far (F) Very Far (VF) The range of the car position to obstacle is from 0 cm to 60 m. Membership function N < 20 m and VF > 50 m. The safe distance of car to the obstacle is 30 m.

There are six step in mamdani fuzzy : • Determining a set of fuzzy rules

978-1-5386-3698-5/17/$31.00 ©2017 IEEE 130

[ABS-57]


Fig 4 Membership Function of Car Position Variable Fig 4 Membership Function of Brake Force Variable Membership function of car velocity are variable : Very Slow (VS) Slow (S) Medium (M) Fast (F) Very Fast (VF) The range of the car velocity is from 0 km/h (0 m/s) to 100 km/h (27.7 m/s). Membership function VS < 20 km/h (5.5 m/s) and VF > 60 km/h (16.6 m/s). The safe car velocity is 80 km/h (22.2 m/s).

The rule base of the fuzzy logic is shown at Figure 6. There are 25 rules in this braking system. Fuzzy control rules are characterized by a collection of fuzzy If - Then rules in which the pre conditions and consequence involve linguistic variables a fuzzy rule have two components: IF-part (also referred to as the antecedent) and Then-part (also referred to as the consequent): If Then . In this paper, the rule is if the position of the obstacle is near and the velocity of the vehicle is fast then the brake force is big etc.

Fig 5 Membership Function of Car Velocity Variable Membership function of brake force are : Small (S) Medium (M) Big (B) The maximum brake force is 9263 N.

Fig 6 Rule Base

III. RESULT The objective of this section is simulating the car braking system using mamdani fuzzy logic control. Figure 7 is showing about the correlation input and output fuzzy logic. The force brake is responding to change on its velocity or distance car to the obstacle. It is obvious from the response that at near car position and slow car velocity there is small brake force used, but it will start changing as car velocity increases or distance from obstacles decreases until reaching

978-1-5386-3698-5/17/$31.00 ©2017 IEEE 131


[ABS-57]

the maximum braking force and very short distances.

The Comparison of Brake Force on Time 0

-100

-200

B rak e Forc e (N)

-300

-400

-500

-600

-700

-800

-900

-1000

0

1

2

3

4

5 Time (s)

6

7

8

9

10

9

10

Fig 9 Braking Force vs Time Graph The Comparison of Veloity on Time 18 16 14 12 V e lo c i t y (m / s )

Fig 7 Surface Viewer

10 8 6 4 2 0

0

1

2

3

4

5 Time (s)

6

7

8

Fig 10 Car Velocity vs Time Graph Initial car velocity is 2.7 m/s then reduce to 0 m/s due to the braking force applied to the car during 2,5s until 3.7s as shown at Figure 9 and 10. The car approaching to the obstacle so it brakes slowly and will stop at the distance 30 m from obstacle. IV. CONCLUSION Designing of a reliable and efficient control system for an car braking system is one of the most challenging issues for automotive. In this paper it have been presented an efficient and reliable design of car braking system using mamdani fuzzy logic control. The result is smooth braking system. The car braking system can started to brake the car 2.5 s with 9263 N until 3.7 s. The car braking system is consist of two inputs and one output The inputs are car position and ar velocity. Each have five membership functions. The output is brake force. It have three membership functions. This control system have twenty five rules. For the future development, the car braking system can be enhanced by appliying more input and membership function so the rule base is more increase.

Fig 8 Rule Viewer Figure 8 is showing the value of car position, car velocity and brake force using mamdani logic control. It showns that if car position is 44.3 m and car velocity is 71.4 km/h (19.8 m/s) then the force used to brake the car is 6170 N.

ACKNOWLEDGMENT The authors gratefully acknowledge partial funding support from the USAID under SHERA program.

978-1-5386-3698-5/17/$31.00 ©2017 IEEE 132


[ABS-57]

[4]

REFERENCES [1] [2]

[3]

National Transportation Safety Comitte Indonesia. 2016. Mamat, M., Ghani, N.M., “Fuzzy logic controller on automated car braking system”, IEEE International Conference on Control and Automation, 2009. Wang, X., Cheng, et.al., “Fuzzy logic controller for electric vehicle braking strategy,” PEDS2007, IEEE, pp. 1542-1547, 2007.

[5]

978-1-5386-3698-5/17/$31.00 ©2017 IEEE 133

Rane, Shubhechchha, et.al., “Performance of Mamdani Fuzzy Inference System for Tracking Multiple Targets Using Autopilot System”, International Journal of Electrical and Electronics Vol. 5, Issue 1, pp: (34-42), 2017. Yadav, Digvijay K., “Modeling an intelligent controller for anti-lock braking system”, International Journal of Technial Research and Application, Vol 3, pp : 122-126, 2015.