2017 2 nd International Conference on Artificial Intelligence and Engineering Applications (AIEA 2017) ISBN: 978-1-60595-485-1 Regulated Voltage Simulation of On-board DC Micro Grid Based on ADRC Technology LUMING CHEN, HAILIANG XU, ZILI LIAO and CHUNGUANG LIU ABSTRACT Output voltage is an important index of on-board DC micro grid, and its stability is affected by many aspects. In the process of power generation, there exist multiple disturbances like speed fluctuations, load changes and time-varying parameters. If various influence factors are not effectively compensated, it will reduce the effectiveness of the system simulation and control precision of practical application. To deal with the problem, an ADRC was applied to on-board DC micro grid and simulated under severe interference conditions. Simulation results based on ADRC were compared with those of the traditional PID controller. The research shows that on-board DC micro grid using ADRC can compensate external disturbances effectively, and has a highly regulated voltage output performance, which verifies the feasibility of the proposed method. KEYWORDS ADRC, PID controller, PMSM, regulated voltage. INTRODUCTION With the improvement of vehicle electrification, kinds of power loads have put forward a higher requirement for power supply capability and quality. Developing onboard DC micro grid becomes an effective means to deal with electrification challenges [1]. On-board DC micro grid transforms mechanical energy into electric energy, and tries to output a relatively regulated voltage for electric accessories. But due to the complicated working environment, unknown disturbances will adversely affect the stability of output voltage and limit its application in high power electric drive systems. In order to solve this problem, researchers have done a lot of works and proposed a variety of different voltage stabilizing methods. Traditional on-board DC micro grid is unusually based on PID control, which is suitable for linear control systems, but it has great limitations in the face of nonlinear system with high quality voltage control requirements [2]. Luming Chen, Department of Control Engineering, the Academy of Army Armored Force, Beijing 100072, China; 18211077415@163.com Hailiang Xu, Zili Liao, Chunguang Liu, Department of Control Engineering, the Academy of Army Armored Force, Beijing 100072, and China. 845
Engine PMSM Rectifier/ Inverter converter Phase current detection Power detection Battery Supercapa citor Loads Rotator position detection Controller Temperature detection Speed detection Charge current detection Reference signal Accelerator position Temperature detection Figure 1. System structure of on-board DC micro grid. Auto Disturbance Rejection Controller (ADRC) is a kind of nonlinear control technology. It does not have high requirements on the accuracy of the model, and can effectively compensate the disturbance. Combined with modern control theories, the method can effectively improve the stability and robustness of the control system when parameters change or disturbances exist [3]. The paper takes on-board DC micro grid of a novel electric armored vehicle as the study object, and apply ADRC technology to improve the system control effect under disturbance conditions. CHARACTERISTIC ANALYSIS OF ON-BOARD DC MICROGRID System Structure On-board DC micro grid consists of a permanent magnet synchronous motor (PMSM), controller, rectifier power converter, rotor position detector and a storage element (Fig.1). Fig.1 shows the controller receives multiplex detection signals from PMSM, power battery, supercapacitor and engine, and its data is processed based on driver s intention. Rectifier / Inverter power converter receives signals from controller and completes the rectifier from three-phase AC to DC as well as the inverted transformation from DC to three-phase AC. Rotor position and speed signal are detected by a built-in rotating transformer and feedbacks are passed to the controller. Another significant element, energy storage device, is powered by battery and supercapacitor, which can meet load demands in both energy and power aspects. PMSM Modeling PMSM mathematical model is usually built in d-q coordinate system, and its internal and external working characteristics can be described from four aspects, namely voltage, flux linkage, torque and motion equation [4]. Its expression is shown as follows: Stator voltage equation ud Rs 0id d q = p + e u q 0 R i s q q d (1) 846
Stator flux linkage equation L 0i 1 d s d = f q 0 L i s q 0 (2) Electromagnetic torque equation 3 Te pn[ diq qid] (3) 2 Mechanical Motion equation dm Te TL J + Bm (4) dt Power Generation Strategy To meet efficient power generation requirements, vector control is usually used to accomplish the stator current decoupling [5]. In terms of structure, the output voltage of the motor is used as the control quantity, the difference between the given voltage and the feedback voltage is used as the regulating amount, and the control function is acted on the control object. The principle of the generation control is shown in figure 2. ADRC DESIGN ADRC Structure As a robust control technique, ADRC combines the advantages of traditional PID controller and the modern control theory, and automatically estimates and compensates dynamic model errors and unknown disturbance [6]. The linear integral series structure has a unique advantage in improving the stability and robustness of the control system. ADRC is constructed from a tracking differentiator (TD), an extended state observer (ESO) and a nonlinear state error feedback (NLSEF). ADRC structure is shown as follows: Reference voltage Voltage controller Current loop Rectifier PMSM Feedback voltage Figure 2. Control schematic diagram of on-board DC micro grid. 847
* v v 1 e U 0 U 0 U U z 2 bu z 1 z 2 bu v Figure 3. The ADRC structure diagram. In the above figure, TD is mainly used for giving an input overshoot to the fast tracking system, and arranging the transition process reasonably to get differential signal. ESO is the core part of ADRC, mainly used for getting all output derivatives and evaluating total disturbances of the observation system. The role of NLSEF is to nonlinearly combine generalized error signal with the disturbance evaluated by ESO, and then ensures that corresponding compensation is available. ADRC Mathematical Model As on-board DC micro grid has been simplified to a first-order system, the tracking differentiator is set as a one order link, the extended state observer is set as a two order system, and the increased order is designed to estimate the total disturbance of system [7]. The mathematical models of each component are shown as follows: Tracking differentiator equation v fal( v v, r, h) (5) 1 1 Extended state observer equation e z1 y, z 1 z2 1fal(, e 1, ) bu, z 2 2fal(, e 2, ), (6) Nonlinear state error feedback equation e1 v1 z1, u0 fal( e1,, ), (7) Nonlinear filter function equation e sgn( e), e ; fal(, e, ) 1 e/, e (8) 848
ADRC Simulation Model Based on ADRC mathematical model, ADRC simulation model can be built and packaged into a full functional module in MATLAB/Simulink. And then, the traditional PID controller model of on-board DC microgrid is replaced by the established ADRC model (Fig. 4). SIMULATION TEST RESULT Before the simulation test of on-board DC microgrid model, it was necessary to adjust parameters of main components and control units. To obtain a relatively accurate simulation mode, PMSM parameters filled into the simulation model (as shown in Table 1); ADRC parameters was adjusted to establish a power control system with high efficiency; finally, according to the given disturbance conditions, the actual test result of ADRC and traditional PID controller was compared. Figure 4. The structure diagram of ADRC simulation model. TABLE 1. MAIN PARAMETERS OF PMSM. Reference index Values Rated power / kw 350 Rated speed / rpm 1710 Stator resistance / mω 7.5 Direct axis inductance / mh 0.8 Quadrature axis inductance / mh 1.5 Rotary inertia / (kg m 2 ) 0.008 Number of pole-pairs 4 Rated torque / Nm 1955 Rotor linkage / Wb 0.538 The ADRC model contains many parameters, which need to set according to the actual regulation experience. Considering the filter effect and the transition process, 849
TD speed factor was set for r =100 and filtering factor was set for h =0.1; parameters of ESO filtering function fal() is set for 1 =05, 2 =0.25, and =0.05; disturbance 1 compensation parameter b=1.5 np f( JLq) was set as the reference value, whose parameters were set for 1 =120, 2 =2500; as for NLSEF, there are parameters of =0.1, =80. To verify ADRC effect, comparative study of ADRC controller and traditional PID controller was tested in Matlab/Simulink, the simulation test results were shown in figure 5. n/rpm (a) Speed disturbance change curve V/V (b) Output voltage change curve T/Nm (c) Motor torque change curve I/A (d) Motor current change curve Figure 5. contrastive analysis of different controller under speed disturbance. 850
At the beginning, the voltage of traditional PID controller rose rapidly with some overshoot, and could reach 750V stable state about 0.6s later. While ADRC voltage rose in a relatively slow speed with no overshoot, it could reach a stable state 0.3s later. At 1.5s, there existed a speed disturbance, the voltage of traditional PID controller got a rapid rise after a first drop and reached a stable state 0.4s later, while the ADRC got a smooth rise after a first droop and reached a stable state 0.2s later. At 3s, the speed jumped from 1400 rpm to 1900 rpm, the former got a rapid decrease after a first rise, and reached steady state 0.5s later, while the latter got a smooth decrease after a first rise, and reached steady state 0.2s later. As a whole, ADRC could effectively solve the contradiction between rapidity and overshoot, and outputted rapid and stable voltage when the overshoot was avoided. SUMMARY According to a certain type of on-board DC micro grid, a simulation model was established based on MATLAB/Simulink, and an ADRC was designed and packaged into the model. Then simulation test results of ADRC and traditional PID controller were compared to verify the effectiveness of the former under disturbance environment. However, this paper mainly focused on the external disturbance but ignored the internal ones. Next, a more comprehensive evaluation system of internal and external disturbance factors will be researched based on present research, aiming to optimize structures and parameters of ADRC and promote regulated voltage characteristics. REFERENCES 1. Liao Zili, Ma Xiaojun, Zang Kemao. Research on status quo and key technologies of all-electric combat vehicle [J]. Fire Control Command Control. 2008, 33(5):1-4. 2. Zhang Xi, Mi Chunting. Vehicle power management: modeling, control and optimization [M]. China Machine Press, Beijing. 2013: 127-138. 3. Zhu Wuxi, Sun Liqing. Control strategy research for extended-range electric bus [J]. Automobile Technology. 2013, 4: 1-5. 4. Xia Yuanqing, Fu Mengyin. Overview of ADRC [J]. Lecture Notes in Control and Information Sciences, 2013, 438(1): 49-54. 5. Tang Renyuan. Modern Permanent Magnet Synchronous Motor theory and design [M]. China Machine Press, Beijing. 2010: 252-262. 6. Wang Zhenglin, Wang Shengkai, Chen Guoshun. MATLAB/Simulink and control system simulation [M]. Electronic Industry Press, Beijing. 2012 :44-64. 7. Xia Changliang, Yu Wei, Li Zhiqiang. ADRC of permanent brushless electric machine torque disturbance [J]. Proceedings of the CSEE. 2016, 26(24): 137-142. 851