Fuzzy-PID in BLDC Motor Speed Control Using MATLAB/Simulink

Brushless DC motors (BLDC) are one of the most widely used types of DC motors, both in the industrial and automotive fields. BLDC motor was chosen because it has many advantages over other types of electric motors. However, in its application in the market, most of the control systems used in BLDC motors still use conventional controls. This conventional method is easy and simple to apply, but has many weaknesses, one of it is that if the system state changes, then the parameters of the PID must also be changed, so that static and dynamic performance will decrease, causing slow response and frequent oscillations. In this study, the design and simulation of a speed control system for BLDC motors using the Fuzzy-PID method was carried out. The research method is performed through simulation with Matlab / Simulink. The simulation is carried out by providing a speed set point of 650 rpm and used two methods, namely Fuzzy-PID and PI. The test results show that the FuzzyPID control can provide better and more stable performance than the conventional PI control. The use of Fuzzy-PID control can reduce speed fluctuation and give torque stability. In the stator current and back EMF (Electromotive Force), Fuzzy-PID has more stable current and voltage than PID which make smaller losses. Furthermore, the proposed method, Fuzzy-PID has faster processing time compared to PID by 6.27%. Keywords—BLDC Motor, PID, Fuzzy-PID


I. INTRODUCTION
Brushless DC motors (BLDC) are among the most widely used DC motors, both in the industrial and automotive fields. BLDC motors are like and consequently called Permanent Magnet DC Synchronous motors [1]. This type of DC motor does not have a brush and commutator, so BLDC motors require less maintenance and can operate quieter than DC motors [2]. The initial cost of the motor may seem expensive but generally diminishes with time [3]. Due to the special characteristic, the BLDC motor is very common in various industrial and biomedical, and robotic applications which require high torque to weight ratio and precise position control for accuracy [4]. BLDC motor was chosen because it has the characteristics of high efficiency, reliability, wide speed range, large torque, and small power [5] [2]. Therefore, it has great theoretical and practical significance to do a lot of research about the brushless DC Motor structure and its controller [6] [7].
There are some different control schemes used for the BLDC motor speed control [8]. The control system on a BLDC motor is quite complicated by using several electronic components that are arranged to be able to switch between the three motor phases with precision [9]. There are two types of speed control systems, namely the Open-Loop system and the Closed-Loop system. An open Loop system is a control system whose output has no effect on controlling the action and there is no feedback [10]. On the other side, the BLDC motor control system on the market already uses a closedloop control system that has a feedback mechanism.
In the market application, most control method in BLDC motors still use conventional controls such as PI (Proportional-Integral) and PID (Proportional-Integral-Derivative) [4] [11]. This conventional control method was chosen because of its simple structure and easy operation [9]. However, this control method still has many weaknesses, such as if the system state changes, the PID parameters must also be changed, so that the static and dynamic performance will decrease, causing slow response and frequent oscillations [12]. Several studies have been developed to produce a reliable and optimal BLDC motor control system. One of the control methods used is fuzzy logic control. This method was chosen because it has the reliability to solve complex and nonlinear problems, is flexible to various problems, and can be combined with other control methods to produce a more optimal system [5].
This research will design and simulate a speed control system for BLDC motors using the Fuzzy-PID method. This test is intended to determine the performance of using Artificial Intelligence, especially Fuzzy Logic in its application in electric motors. Fuzzy logic is used to generate PID parameters (Kp, Ki, and Kd) which are then forwarded to the PID to produce a more reliable and optimal control system. The test results are then compared using the conventional PI method in order to obtain which method is best used to control the BLDC motor speed.

A. BLDC Motor
The BLDC motor is a type of permanent magnet synchronous motor, which has a permanent magnet in the rotor and a trapezoidal back-electromotive force (EMF) [13]. BLDC motor has the characteristics of a DC machine by replacing the mechanical commutator and brush with a solidstate switch and there is no electrical connection between the stator and the rotor [14]. BLDC motor construction is shown in Fig. 1.
BLDC motors are widely used in various electronic components, especially in electric vehicles. This is because BLDC motors have various advantages over other motors, namely wide torque range, high speed, high efficiency, good Journal of Robotics and Control (JRC) ISSN: 2715-5072 9 Hari Maghfiroh, Fuzzy-PID in BLDC Motor Speed Control Using MATLAB/Simulink dynamic response, strong, no-slip, and others [14]. The motor specification that used in this research is described in Table  1.  The electrical part of the system is governed by equation (1) [15]: The mechanical part of the motor is governed by torque equation as:

B. Various BLDC Control Scheme
Several studies have been conducted to combine the Fuzzy control system with other control systems for application in BLDC motor drives. Some of the studies that have been carried out are as resumed in Table 2.

C. Fuzzy Logic Controller
Fuzzy Logic is a branch of Artificial Intelligence (AI) that has been used since 1965 until now. Fuzzy is still chosen because of its reliability to solve complex and nonlinear problems, its flexibility to various problems and can be combined with other control methods to produce a more optimal system [5]. Fuzzy logic uses basic rules to produce fuzzy output, namely the IF-THEN rule, where IF is an antecedent and THEN is a consequence [31]. In the fuzzy method, there are 4 main components, namely [5]:  Fuzzifier: Fuzzifier is used to map the value / price of variables in the real world into fuzzy sets.  Knowledge base: The knowledge base contains control system knowledge as a guide for evaluating the state of the system to obtain control output as desired by the designer.  Fuzzy inference engine: Fuzzy inference engine translates fuzzy statements in the rule base into mathematical calculations (fuzzy combinational).  De-Fuzzification: Defuzzification can be defined as the process of changing Fuzzy quantities which are presented in the form of output Fuzzy sets with a membership function to regain their crisp form.
The block diagram of the Fuzzy control system can be observed in Fig. 2.

D. Fuzzy-PID Block Diagram
The block diagram of the Fuzzy-PID control can be seen in Fig. 3. In this structure, 3 Fuzzy Logic Control blocks are used, each of them has the same input, namely error and delta-error. Meanwhile, the output of each fuzzy block is Kp, Ki and Kd. This output is then processed using the PID control to adjust the output voltage from the power source that supplies the BLDC motor.
Since there is five fuzzy membership for each input; therefore, the fuzzy rules used are 25 rules. These rules can be seen in Table 3. Where DB: Decrease Big; DS: Decrease Small; NC: No Change; IS: Increase Small and IB: Increase Big. While the set of membership functions of the input variables (error and delta error) can be observed in Fig. 4 and Fig. 5.  The Fuzzy block diagram for output of Kp, Ki and Kd are shown in Fig.6, Fig.8 and Fig. 10, respectively. Whereas, Fig.7, Fig.9 and Fig. 11 shows the Membership function of output Kp, Ki and Kd respectively.

III. RESULT AND DISCUSSION
The simulation is carried out by providing a speed set point at 650 rpm. The simulation was carried out by comparing two methods, namely Fuzzy-PID and the conventional PI method. The plant of this circuit is a brushless DC motor (BLDC) 48V 1kW with a maximum speed of 700 rpm. Fig. 12 shows the simulation block of the proposed Fuzzy-PID control method in MATLAB Simulink. Fig. 11. Membership function of Kd output Based on Fig. 13, it can be observed that the Fuzzy-PID control is able to provide a good response to reach the desired set point. By using the Fuzzy-PID method, steady-state speeds can be achieved at 0.3 seconds without significant overshoot and fluctuation. On the other side, the conventional PI method can reach a steady-state point at 0.2 seconds but with very high fluctuations and experiencing overshoot up to 180 rpm.
In the electromagnetic torque, Fig. 14 shows that the Fuzzy-PID method produces a stable torque graph, where the torque decreases with the increasing speed of the motor. The torque reaches a constant point at 5 Nm when the motor reaches a steady-state speed. While the PI method provides an unstable graph of torque with high enough fluctuation and overshoots up to 35 Nm before the steady-state speed is reached. Fig. 15 shows that the current and back voltage resulted from Fuzzy-PID is quite stable during the motor operation. In contrast, when using the PI method, there are fluctuations in the current and reverse voltage EMF on the stator. This resulted in system instability and resulted in a lot of losses. The simulation time for each method is resumed in Table  IV. It clearly shown that Fuzzy-PID has faster processing time compared to PID by 6.27%. Therefore, the overall result show that Fuzzy-PID is better than PID.

IV. CONCLUSIONS
The design and simulation of the Fuzzy-PID control system on BLDC speed control have been successfully carried out. Based on the simulation, the Fuzzy-PID control can provide better and more stable speed performance than using the conventional PI control. The use of Fuzzy-PID control can reduce speed fluctuation and give better torque stability so that the BLDC motor can operate more reliably. In the stator current and back EMF, Fuzzy-PID has more stable current and voltage than PID which make smaller losses. Furthermore, the proposed method, Fuzzy-PID has faster processing time compared to PID by 6.27%.