TMCnet News
The Research on the Design and Fuzzy Control of Single Wheel Intelligent Self-balancing Scooter [Sensors & Transducers (Canada)](Sensors & Transducers (Canada) Via Acquire Media NewsEdge) Abstract: The research of inverted pendulum system can effectively reflect the control of many typical problems, such as nonlinear problem, robustness, calm, follow-up and tracking problems. Control the unicycle smooth running principle is similar to inverted pendulum principle, whose core design thought is through the single round the sensor real-time measuring unicycle related information. Based on the analysis of the design concept and design flow, this article first designs the Foot pedal design, Side plate design and determined Angle. At the same time it chooses motor and battery to do the in-depth research. And then it tests the performance of the motor and gets the scooter assembly drawing. After that it studied sensor and MCU. The article chooses the most match themes of car performance and requirements plan. According to people's normal thinking mode and life experience, the article constructs the control model. The article determines appropriate membership, makes proper fuzzy rules and fuzzy domain, finds out input and output domain and figures out the simulation debugging. At the same time it also completes the motor remote control debugging, realizes the design and related functions. To a certain extent, it solves the Environmental problems, the traffic problems and energy problems which are increasingly prominent today. Copyright © 2013 IFSA. Keywords: Inverted pendulum system, Intelligent self-balancing scooter, Fuzzy control research. (ProQuest: ... denotes formulae omitted.) 1. Introduction Environmental problems, traffic problems and energy problems are increasingly prominent today, environmental protection, energy saving and safe comfortable walking tools become the main themes of modem time. In order to fill the blank of the research on single wheel self balancing car, the article makes the study on the design and fuzzy control of single wheel intelligent self-balancing scooter [2]. The research of Inverted pendulum self-balancing car began in 1987, started by a professor of the Japanese telecommunications university who put forward a similar design thought. At present, many of the domestic and foreign universities and research institutions have self-balancing electric vehicles to do the theoretical research, prototype making and experiments. Some of them have made some achievements [4]. However, it mainly concentrated on robots or double self-balancing walking electric cars. Research on unicycle is just at the initial stage. As this research on single wheel intelligent self-balancing skateboards car is the first national research of the country, the research will motivate the development of the agitation of national single wheel intelligent self-balancing scooter. Inverted pendulum control system is a complicated and unstable, nonlinear system, which purpose is to control theory teaching and carry out various control experiments in ideal experimental platforms [5-7]. The research of inverted pendulum system can effectively reflect the control in many typical problems, such as nonlinear problem, robustness, calm, follow-up and tracking problem and so on [1]. Through the inverted pendulum control, it can be used to test new control method, which have stronger ability to deal with nonlinear and instability problems [6]. Inverted pendulum control problem is to make pendulum rod as early as possible to achieve a balance position, and there was no big oscillation and big Angle and speed. When the pendulum lever got expected position, the system can overcome random disturbance and maintain a stable position [3]. Inverted pendulum system is input for car displacement (namely position) and pendulum lever Angle of expectations, the computer in every sampling cycle collection from sensor car and pendulum bar of the actual position signal, and expectation comparison, through the control algorithm to control the quantity, then through digital-to-analog driven dc motor to realize real-time control of the inverted pendulum. Dc motor through the belt drives the car in a fixed orbit movement, swings rod end installed in car. This point can be used as the axis to make pendulum rod in the vertical plane swing freely. The control principles are shown in Fig. 1. Force U parallel to the direction of the track effect on car and make the bar around the axis of the car in the vertical plane internal rotation, the car would go along the horizontal movement track. When there is no force, pendulum rod would be in a vertical stable equilibrium position (vertical down). In order to make the pole swing or achieve vertical upward stability, it is necessary to give the car a control, make it in orbit to be forward or toward the rear. Control the unicycle smooth running principle which is similar to inverted pendulum principle, its core design thought is through the single round the sensor real-time measuring unicycle related information (such as unicycle's Angle and speed, etc.) [8]. Then it will transfer the information to controller. After processing these signals, the control system would adjust the motor speed, motor with the appropriate torque and speed drive wheel running, to ensure unicycle balance and safety driving [10]. At the same time, the unicycle would be considered. This research mainly studies the inverted pendulum control, which is the basic principle of designing reasonable unicycle mechanical structure and control system. Finally, the research can realize unicycle simple control and safe driving. And we use an adult standard of height and weight as unicycle mechanical-design criterion, which should satisfy the load carrying capacity for 70 kg - 80 kg requirements. In practical operation, the operator only through the body's center of gravity offset to adjust the board Angle, so as to change the unicycle of the speed and deflect the directions. 2. Hardware Analysis Study and the Overall Design of Self-balancing Scooter This design not only fusions the car industry in the development of the three subjects, but also better than the existing modem cars, which has more extensive applicable line. Even though it is narrow space range can also free flexible motion, realize the short distance manned driving. At the same time, for the clever application of inverted pendulum principle, the design is unique. 2.1. Mechanical Structure Design 2.1.1. The Design Concept and Design Flow According to the demand and market analysis, the following process can be determined: 1) To determine whether the design requirements and custom machine scheme. 2) The mechanical system is divided into several modules: Shell module, Radius module, Hub motor module, Battery module, SCM, sensors, wire, controller module, etc. 3) For each function module in concrete structure and dimension design, draw the part drawing and the whole assembly drawing. 4) The system to conduct a comprehensive check and determine the material. 5) Mechanical parts processing. 6) Assembly and debugging. 2.1.2. Foot Pedal Design, Side Plate Design and Angle to Determine 1) Foot pedal design. According to one foot length and the average person's weight value to design Mmax (refer to the man-machine engineering standards of the adult average data).The size is 250 mm x 173 mm x 3 mm sheet metal and 90 kg load, and mainly is bending deformation. Force analysis model can be seen in Fig. 2. h=3 mm, b= 173 mm, F=450 N. Mmax = Fxb/2 smax = Mmax /w = 18.87Mpa < [s] = 160Mpa ... smax = Mmax = 99.35Mpa < [s] = 100Mpa Through this model, it uses the material mechanics theorem, intensity, select materials for 45 steel. 2) Side plate design. The Size is 700 mm x 130 mm x 3 mm sheet metal, force analysis model are as follows Fig. 3(a) and (b). F=271 N, a=250 mm, b=3 mm, h = 130 mm, .... Through this model, the use of material mechanics theorem, intensity, and comprehensive consideration of the cost problem, select steel for Q - 235. 3) Angle to determine. The board Angle determination in 30 degrees or so, and then defines hub motor shaft location dimension to ensure this Angle value. 2.2. Motor and Battery Selection Research Through the analysis of the above contents, it can be concluded that the control and load stability is good, and grade according to the design requirements of the 8-10°, without a lot of acceleration shock, and the motor circuit does not appear short circuit and out of control. Then the basic design parameters can be obtained, which can be seen in Table 1. Have the above data for that resistance torque: ... Motor power: W = 8.7*T1 *ohgr;/(1000*?1 *?2). The calculation can be made: W=482.1443J. Battery validation formula for: Imax<W/U/ ?3. Experience card selected motor JN ZWD 48B -081202meet the requirements. Motor test conclusion shown in Table 2. Based on the particularity of the motor, 500 w, 48 v, maximum allowable current we ask for more than 15 A, we have the choice of 12 ah, 12 v, 2.6 kg, 14 A maximum current, We bought four block meet motor requirements. 2.3. The Overall Size and Assembly Drawing Through the above analysis, the article designs the article draw body overall dimensions, which can be seen in Table 3. Here Fig. 4 is to design the overall appearance of diagram. 3. The Sensor Characteristics and SCM Choose Research 3.1. Sensor Research Analysis of single wheel balance from the function of the car, it can be seen that the project need two sensors: measuring the car angular velocity gyroscope and the accelerometer of the car acceleration measurement. Reference to the similar product option and based on the practical needs, the choice schemes can be got in the following: 1) ADXRS300 type of single angular velocity gyroscope and ADXL203 type of biaxial Angle accelerometer, ADXRS300 output voltage and the relationship between the measured angular velocity below (set by measurement of the angular velocity for av, the unit is ° / s; output voltage for U0, the unit is mV; sensitivity K for 5 mV/ °/ s, zero output voltage of 2.5 V. U0 -250O=K x av [9]. ADXL203 is the solid structure of crystal silicon of the accelerometer. Under the influence of the temperature of minimal, it can ensure the stability of the system. As its volume is small, the biaxial input can simplify circuit design and installation. However, gyroscope and Angle accelerometer separate scheme needs to spend a lot of energy time to connect between the sensors circuit. Thus, the debugging is relatively complicated. 2) The United States ADI produced new product ADIS 16300, which is a built-in yaw angular velocity gyroscope and accelerometer with three axes in the integrity of the inertial system, which can meet the requirements on measuring the accuracy. And complicated discrete design scheme, comparing to ADIS 16300 for accurate shaft inertia detection and industry system integration, provides a simple and cost-effective method. All necessary movement test and calibration are factory production part of the process, greatly reducing the integration time of system. Improved SPI interface and register structure provides a faster data collection and configuration control. ADIS 16300 and flexible interface suitable for use ADIS1635X series products of the current system, only need four degrees of freedom for the inertia test system, which provides low cost solutions. ADIS 163 00 can provide optimal dynamic performance, but the price is expensive [12]. 3) The factors that influence the selection sensor have the size of the working voltage, the accuracy of measurement, measurement range, etc. By analyzing the final use models, the article chooses SCAIOOT - D01 biaxial Angle sensor. A) SCA100T-D01 sensor characteristics. This type sensor for biaxial measurement (X and Y), measuring range is±0.5g (plus or minus 30°), according to the movement condition of the car, it can be seen that the Angle is enough. Its sensitivity is 4v/g, nonlinearity is ±2 mg, for sensing element control frequency rate response, using 12 feet sealants SMD package, has the superior internal and external failure detection function. Using 5V single voltage provide, and power supply electric voltage of a linear relationship between the output, with SPI digital output and 0.5 v - 4.5 v analog signal output, its built-in temperature sensor for temperature compensation and temperature signal output. This type sensor is compared with other sensors, its advantages are: 1) In the temperature and time of high reliability and stability, 2) Instrument stage performance, 3) A wide range of temperature range, 4) High load and high resistance to shock, 5) Low noise and high resolution in accordance with the above data, combined with the unicycle movement characteristic and measurement requirements, so that the sensors have already had. Have the need to perform, it can accomplish Angle measurement, like Fig. 5 and Fig. 6. B) Using sensors This project used this Angle sensor to measure the Angle. The sensor pin schematic diagram and measurement methods can be seen in Fig. 7 and Fig. 8. C) Sensor debugging This sensor needs to supply 5 V dc regulated power supply, which can output analog signal, and then the MCU AD conversion function, into digital signal. Sensor part debug program: #include <AT89x52.h> #include <intrins.h> #defme LED Pl_5 sfrADC_CONTR=OxBC; sfr ADC_RES=OxOBD; sfr PlASF=0x9D; sfr ADC_RESL=OxOBE; sfr AUXRl=0xA2; void delay(unsigned int n) {while(n!=0){_nop_0;n-;}} void main(void) { IE=0xA0; AUXRl=0x00; PlASF=0x04; ADC CONTR |= 0x80; delay(5); ADC CONTR |= 0x02; ADC CONTR |= 0x08; while(l){;}} 3.2. SCM Choose Research Based on the SC performance and movement characteristics of consideration, the article chooses the SCM model for one-chip computer C8051F120. It is fully integrating mixed signal chip system of the type of MCU chip. Specific parameters and characteristics are as follows: 1) The kernel for high speed, pipeline structure of CIP-51, is fully compatible with classical C51, top executive rate for 100 MIPS. 2) At full speed, noninvasive online debugging interface (JTAG), using USB interface simulator. 3) 128 k FLASH storage space. 4) 8448 (8 k + 256) bytes of XRAM, can expand external data memory, 64 k addressing space. 5) 64 (8 x 8) I/O pins. 6) Two UART serial interfaces. 7) There are 12 D/A and A/D, timer, capture/comparison module, the programmable counter/timer, the watchdog timer, VDD monitor and temperature sensor, support SPI, SMBUS/the I2C, etc. 4. Fuzzy Control, Simulation Research and Motor Remote Debugging Research of the Self-balancing Scooter The article adopts the fuzzy control to control the car balance. The main advantages are as follows [11]: does not need to build a very accurate mathematical model, can suit people's normal thinking mode and life experience in constructing control model. The difficulty lies in the division of the reasonable theory field to determine appropriate membership and make proper fuzzy rules. Control principle diagram can be seen in Fig. 9. 4.1. Fuzzy Control Domain Division Domain division and the structure of control system need to make sure that the input and output quantity [13]. With Angle and angular velocity for input quantity, the motor drives force for output is shown in Fig. 10. 1) The input output domain. Input quantity 1, as Fig. 11: Angle, its scope as [-0.3, 0.3] between radian, specific subdivided into: NL, NM, NS, ZE, PS, PM, PL seven domains. In order to guarantee in the equilibrium position control accuracy, at 0 degrees or so divided universe of the division of the fine point. Input volume 2, as Fig. 12: angular velocity, its scope as between [-15rad/s, 15rad/s], Specific subdivided into: NL, NM, NS, ZE, PS, PM, PL seven domains. With Angle domain division, in order to guarantee in the equilibrium position control accuracy, at 0 degrees or so divided universe of the division of the fine point. 2) Output, as Fig. 13: motor driving force. Its scope as [-12n, 12n] between, specific subdivided into: NL, NM, NS, ZE, PS, PM, PL seven domains. 4.2. The Fuzzy Rule and the Simulation Debugging 1) The fuzzy rule. The formulation of fuzzy rules is based on experience to carry out. For example, slide Angle is big, board face Angle velocity is large and the motor driving force will be big. This analogy, work out the input and output quantity of the corresponding relationship between fuzzy, seen as Fig. 14 and Fig. 15 show: 2) The simulation. Using the Matlab Smulink simulation module, it constructs a good control model, and then the simulation, is very convenient. The Fuzzy Logic Controller is after encapsulation of Fuzzy control module, the internal is as bellows: it is possible to construct the Fuzzy control model introduction to this module, can be simulated. Finally, it can get the simulation animation, which can be seen in Fig. 16. 3) Debugging. Through the computer simulation, fuzzy control model for debugging is obtained. Debugging principle is as follows: the first from a small dip balance, and gradually expands to large Angle balance, and constantly test, unceasingly carries on the membership degree and fuzzy rules of the adjustment, which makes the car finally achieve stable equilibrium state. Although the simulation is in the computer control system, virtual simulation and physical simulation are very different. For example, it did not consider friction, sports vibration and the influence of such factors as temperature, which is needed in physical debugging. All parts together, and the sensor properly installed in car body, were to ensure the measurement stability. Physical debugging stage for control program repeated changes in order to make the process to the physical work stably and reliable. Fig. 17 shows the control system circuit connection. 4.3. Motor Remote Debugging Research of the Self-balancing Scooter 1) Welding procedure bum writing board: main function is to connect controller and computer. 2) Welding the SCM minimum system board, including 5V/12V voltage conversion circuit, indication screen display interface, the display lamp, debugging performance is good, making the corresponding input module, and install the remote control module, can now through the remote control into line control motor stalling and positive ¿¿negative. Diagram is the single-chip microcomputer control system, remote control circuit links and remote control switch. 3) Motor controller: able to drive 500w motor. There are four input pins, respectively CO, Cl, PWM, GRAND, give CO and Cl common control motor positive 8c negative, and PWM wave control motor speed. Output voltage is more than lOv, voltage and waveform duty-factor strict linear relations, through a single chip microcomputer output waveform of duty. 4) Display module: can display the movable infinite switch (the left bank light display), program operation instructions (through a luminous two over), as well as the buzzer to display the Angle, which is too big when alarm (need to program to set, convenient to do process sequence classmates debugging). 5. Conclusion Inverted pendulum control system is a complicated and unstable, nonlinear system, which is based on the analysis of the design concept and design flow, designed the Foot pedal design, Side plate design and determined Angle. At the same time the article chose motor and battery to do the in-depth research, and tested the performance of the motor, the research got the scooter assembly drawing. Further the article studied sensor and MCU, in the sensor three schemes the article chose the most match of car performance and requirements plan. According to people's normal thinking mode and life experience, the article constructed the control model, determined appropriate membership, made the proper fuzzy rules and fuzzy domain, found out input output domain, and then figured out the simulation debugging. At the same time it also completed the motor remote control debugging, realized the design and related functions. References [1] . Felix Grasserm, Aldo D'arrigo, Silvio Colombi, Alfred Rufer., JOE: A Mobile, Inverted Pendulum, IEEE Transactions on Industrial Electronics, Vol. 49, No. 4,2002, pp. 107-114. [2] . Zhang Pei Ren, Zhang Zhi Jian, Zheng Xu Dong, Zhang Hua Bin, Based on the 16/32 bit DSP robot control system design and realization, Tsinghua University Press, China, 2006. [3] . Trevor Blackwell, Building a Balancing Scooter, http://www.tlb.0rg/#sc00ter [4] . Trevor Blackwell, Balancing Scooter Version 2, http://www.tlb.Org/#scooter2 [5] . M. A. Ciar, J. B. Field, S. G. McMahon, P. S. Philps, EDGAR, A Self-Balancing Scooter Final Report, The University of Adelaide, Australia, 2005. [6] . N. P. Baker, C. P. Brown, D. R. S Dowling, J. L. Modra, D. J. Tootell, SON of EDGAR Final Report, The University of Adelaide, Australia, 2006. [7] . Brian Beckwith, Eric Desjardins, Chris Howard, Joel Murphy, Matt Uganecz, Jack Woolley, HTV Project Final Report, Camosun College, Canada, 2004. [8] . Cossalter, Vittore, The influence of frame compliance and rider mobility on the scooter stability, Vehicle System Dynamics, Vol. 45, No. 4, April 2007, pp. 313-326. [9] . Sheu, Kuen-Bao, Simulation for the analysis of a hybrid electric scooter powertrain, Applied Energy, Vol. 85, No. 7, July 2008, pp. 589-606. [10] . Müller, Loretta, Investigating the potential for different scooter and car exhaust emissions to cause cytotoxic and (pro-)inflammatory responses to a 3D in vitro model of the human epithelial airway, Toxicological and Environmental Chemistry, Vol. 94, No. 1, January 2012, pp. 164-180. [11] . Lee, Yuang-Shung, Kuo, Tsung-Yuan, Wang, Wei-Yen, Fuzzy neural network genetic approach to design the SOC estimator for battery powered electric scooter, in Proceedings of the IEEE Annual Power Electronics Specialists Conference PESC Record, Vol. 4, No. 3,2004, pp. 2759-2765. [12] . Aoshima, Ichiro, Development of electric scooter driven by sensorless motor using D-state-observer, World Electric Vehicle Journal, Vol. 3, No. 1, 2009, pp.125-130. [13] . Frendo, F., Simulation of motor-scooter drop testing by a multi-body finite element integrated approach, Proceedings of the Institution of Mechanical Engineers, Vol. 219, No. 4, December 2005, pp. 383-391. 1 Gao Xiaoxing,2 Li Xiaoxia,1 Cui Han 1 College of vocational and technical, Shijiazhuang University of economics, Hebei, Shijiazhuang 050031, China 2 The Modem Education Centre, Shijiazhuang University of economics, Hebei, Shijiazhuang 050031, China Received: 16 December 2013 /Accepted: 29 December 2013 /Published: 30 December 2013 (c) 2013 International Frequency Sensor Association |