Abstract: Magnetic levitation brushless DC generators are used in wind power generation systems, offering advantages such as high speed, no lubrication, no wear, no mechanical noise, no need for sealing, high precision, and long lifespan. Magnetic levitation systems are typical nonlinear hysteresis systems, making it difficult to establish accurate mathematical models, and conventional PID control often fails to achieve good control results. To address this challenge, a Fuzzy-PID hierarchical controller was designed for magnetic levitation systems. This controller combines the advantages of fuzzy control (small overshoot, good stability and robustness) with PID control (fast speed and high precision). Furthermore, it is a contactless switching method without fixed thresholds; the fuzzy rule-based switching ensures a smooth transition between the two control methods, thus exhibiting good setpoint tracking capability and strong anti-interference ability. Simulations of the magnetic levitation brushless DC power generation system were conducted using Matlab software. Simulation results show that the Fuzzy-PID hierarchical control enhances the system's anti-interference ability and demonstrates strong robustness. Keywords: wind power generation, magnetic levitation system, brushless DC generator, fuzzy PID hierarchical control 1. Introduction Magnetic levitation bearings, which have been developed in the past 30 years, are a new type of high-performance bearing that uses magnetic force to levitate the rotor in space and achieves no mechanical contact between the rotor and the stator. Magnetic levitation bearings have advantages such as high speed, no lubrication, no wear, no mechanical noise, no need for sealing, high precision and long life. However, due to the axial space and volume weight occupied by the magnetic levitation bearing itself, the low axial utilization rate limits its critical speed and output power, which affects the lightweight and miniaturization of high-speed motors. In addition, magnetic levitation bearings make the cost of motors too high and the dynamic response slow, which affects its widespread use. Taking advantage of the similarity between the magnetic levitation bearing and the motor structure, the magnetic levitation bearing winding that generates magnetic levitation force is placed in the motor stator, eliminating the need for a dedicated magnetic levitation bearing (as shown in Figure 1). By decoupling the torque winding and the levitation force winding [1], the rotor can be stably levitated, so that the motor rotor can generate torque and self-levitate at the same time. In this way, the so-called bearingless motor is formed. Applying this technology to small wind power generation systems will be a major breakthrough in wind power generation system research. Compared with traditional wind power generation systems, magnetic levitation bearingless wind power generation systems have a series of advantages such as small size, high power density, high speed, no lubrication, no wear, no mechanical noise, and long life. They have broad application prospects and have received increasing attention. Magnetic levitation systems are typical nonlinear systems. The control system can only work normally if it meets the requirements of stability, speed, robustness and anti-interference. However, the characteristic parameters of magnetic levitation systems change with time, making it difficult to establish an accurate mathematical model [2]. Therefore, although traditional control can complete the control, it is difficult to achieve ideal control performance. Fuzzy control is a language control that does not depend on the mathematical model of the controlled object. It can be directly optimized from the operator's experience, but the control accuracy is not high and the steady-state performance is poor. A hierarchical control method combining fuzzy control and PID control is adopted. This can ensure that PID control has no steady-state error and good static stability, and also has the characteristics of strong adaptive capability and good dynamic performance of fuzzy control. This paper combines two methods to design a controller: fuzzy control is used when the error is large, enabling the system to approach the setpoint with good dynamic characteristics and small overshoot; when the error is small, i.e., when it tends to stabilize, PID control algorithm is used to leverage the advantages of PID control such as precision and small static error, further improving static characteristics. A contactless switching method is also designed. Simulation and experiments show that this method has good control effect and strong robustness. [align=center] Figure 1 Principle of levitation force generation of magnetic levitation brushless DC generator[/align] 2. Working principle of magnetic levitation brushless DC generator A brushless DC generator (BLDG) consists of a generator body and peripheral circuitry. The generator body is a permanent magnet synchronous motor. Figure 2 shows the principle of magnetic levitation force generation of a bearingless sinusoidal brushless DC generator. In addition to torque windings Na and Nb, the stator also has levitation force windings NA and NB. Here, for ease of analysis, the three-phase torque windings and three-phase levitation force windings of the motor are equivalent to two-phase windings respectively. Assuming the rotor is in the center position, without eccentricity, and the motor is unloaded. A uniform 4-pole magnetic flux is generated by a permanent magnet. Since the motor is unloaded, the current i4 in the torque windings Na and Nb is approximately 0, thus the magnetic field generated by this current can be ignored. When a positive current flows through the NA winding while no current flows through the NB winding, a 2-pole magnetic field of 5A in the direction 'a' is generated, with its magnetic field lines shown in the figure. The combined result of the 2-pole and 4-pole magnetic fields will increase the air gap magnetic flux density at point 2 in the figure, increasing the magnetic field gap force, while decreasing the air gap magnetic flux density at point 1, decreasing the magnetic field gap. Ultimately, this results in a net force F in the negative direction 'a' on the rotor to overcome the load in this direction. If a reverse current flows through Na, the rotor will experience a magnetic levitation force in the positive direction 'a'. Similarly, energizing the NB winding can generate a magnetic levitation force in the direction of the B axis. By adjusting the current in NA and NB, a magnetic levitation force in any direction can be generated to overcome the load in any direction, ensuring stable rotor levitation. 3. Fuzzy-PID hierarchical control algorithm for magnetic levitation BLDG The design and implementation of the control system is the difficulty and key point of magnetic levitation brushless DC generator. The control system consists of two parts: torque control and levitation displacement control. The design of PID controller, high-precision sensing and detection device, and real-time effective control algorithm are the key to the control system. Since the system is high-order, it is difficult to design PID regulator using traditional engineering design methods. This paper adopts the Fuzzy-PID hierarchical control method. The switching of conventional Fuzzy-PID control is based on the pre-given deviation range. The selection of the switching point becomes the key to affecting the system performance [3]. Switching too early will not reflect the advantages of fuzzy control and will increase the overshoot. When switching too late, if the PD type fuzzy controller has a large steady error, it may not be able to enter the PID controller and become a single fuzzy control form. Therefore, it is not easy to achieve a good control effect by selecting the switching point based on experience [4]. In addition, when the two controllers switch, it is difficult to ensure that their output is equal and that the control quantity is continuous and does not jump. Therefore, disturbances are inevitable during the switching process, which increases the overshoot and lengthens the adjustment time. To address these shortcomings, a fuzzy-PID hierarchical controller (FSFC, Fuzzy Switched Fuzzy-PID Controller) based on fuzzy rule switching is designed below. As shown in Figure 2. The main shortcoming of the conventional fuzzy-PID hierarchical controller is that it switches according to a fixed threshold of error magnitude [5]. To solve this problem, a contactless switching method can be designed. The fuzzy-PID controller based on fuzzy rule switching is switched by the following rule: If e is Z[sub]e[/sub] and ec is Z[sub]ec[/sub], then U is U[sub]p[/sub] else U is U[sub]f[/sub] where U[sub]p[/sub] and U[sub]f[/sub] are the outputs of the PID controller and FLC, respectively, and z[sub]e[/sub] and z[sub]ec[/sub] are the membership functions of the fuzzy switching rule, as shown in Figure 3. a and b are the input ranges of error and error change rate, respectively. By changing the values ​​of a and b, control components of different intensities can be obtained. When the input error is ei and the rate of change of the input error is eci, their corresponding membership degrees are u[sub]e[/sub] and u[sub]ec[/sub], respectively. Here, the "and" operation can take the product or the smaller value. Taking the smaller value as an example, the output intensity coefficients of the PID controller and FLC are respectively: Figure 3 Membership function of FSFC fuzzy switching rule 4. Design of Fuzzy-PID hierarchical controller for magnetic levitation BLDG 4.1 Design of specific controller parameters 4.2 Establishment of fuzzy control rules Because the triangular membership function has a high resolution, the fuzzy subsets of E, Ec, and U in this paper all adopt the triangular membership function. Its function curve is shown in Figure 3 (since the function curves of E, Ec, and U are the same, they are only marked in the same curve graph). [align=center] E, Ec, U Figure 3 Membership function curves of fuzzy subsets of E, Ec, and U[/align] This paper sets seven levels according to the model: negative large NB, negative medium NM, negative small NS, zero ZE, positive small PS, positive medium PM, and positive large PB. Based on this, the control rule table is established as follows. [align=center]Table 1 Fuzzy Control Rule Table[/align] According to the above inference rules, the fuzzy reasoning process is completed by using the maximum membership method to resolve the fuzziness. 5. Simulation and Experiment For the above magnetic levitation system, the system is simulated using MATLAB's FUZZY toolbox and SIMULINK simulation environment. The simulation structure diagram is shown in Figure 4. [align=center]Figure 4 Simulation structure diagram of Fuzzy-PID hierarchical control of magnetic levitation BLDG[/align] The simulation results of the system's step response are shown in Figure 5. Compared with traditional PID control, the system's overshoot is significantly reduced, oscillation is significantly reduced, and it has a small steady-state error, ensuring good steady-state accuracy. [align=center]Figure 5 Comparison of Simulation Results[/align] 6. Conclusion Applying a magnetic levitation brushless DC generator to a small wind power generation system is a major breakthrough in wind power generation system research. Given the difficulty in establishing accurate mathematical models for the control process of fuzzy wind power systems, it is challenging to guarantee control effectiveness using only PID control. Fuzzy control, on the other hand, does not require precise modeling of the controlled object. Therefore, combining the two for hierarchical control can complement each other and leverage their respective advantages. Simulation results show that when this hierarchical control method is applied to a magnetic levitation wind power generation system, the system's overshoot and oscillations are significantly reduced, while maintaining good steady-state accuracy, resulting in a relatively good control effect. References [1] Lu Jiankang, Fan Huiyun, Fang Xiaoting, Gao Yang. Decoupling control of magnetic levitation bearingless motor [J]. Micromotors, 2006, 39 (6): 15-18 [2] Zhou Yuan, He Yikang, Nian Heng. Complete system modeling of permanent magnet type bearingless motor [J]. Proceedings of the CSEE, 2006, 26 (4): 134-139 [3] Bong Joo Kim, Chang Choo Chang． Design of fuzzy PD+I controller for tracking control[C]． American Control Conference, Proceedings of the 2002, 2002, 2124-2129． [4] Radu-Emil Precup ＊, Stefan Preitl. PI-Fuzzy controllers for integral plants to ensure robust stability[J]. Information Sciences 177 ,2007, 4410–4429. [5] Lu Huacai Xu Yuetong ]rang Weimin Chen Zichen, Fuzzy PID Controller Design for a Permanent Magnet Linear Synchronous Motor Feeding System[J]. Transactions of China ElectrotechnicaL Society, 2007,22(4):59-63. [6] V. Mukherjee, SP Ghoshal , Intelligent particle swarm optimized fuzzy PID controller for AVR system[J]. Electric Power Systems Research .2007,1689–1698 [align=center]The Research on Fuzzy-PID Controller-FSFC for wind driven- maglev BLDC generator YANG Guo-liang LI Hui-guang (College of Electrical Engineering, Yanshan University, Qinhuangdao 066004, China)[/align] Abstract : Maglev brushless DC generator (BLDG) is adopted in the wind power generation systems (WPGS), which has the advantages of high-speed, no-lubricating, no-wearing, no-mechanical noise, no-sealed, high-degree of accuracy and long life. It is difficult to apply general PID control to the maglev system with the nonlinear and hysteresis characteristics and is difficulty in establishing extract model. In order to solve the problem, this paper designed the Fuzzy Switched Fuzzy—PID Controller (FSFC) for maglev systems． The FSFC combines the advantages of fuzzy control and PID control, such as tiny overshoot, good stability, speedliness, and high precision. The fuzzy switching ensures a calm and stable transition, giving it strong tracking ability and anti-jamming capabilities. Simulation of the Maglev BLDG system in Matlab/Simulink demonstrates that FSFC exhibits excellent control performance in terms of stiffness and robustness in resisting disturbances. Keywords : wind power generation; Maglev; BLDG; FSFC Author Biographies: Yang Guoliang (1973-), male, from Gongzhuling, Jilin Province, PhD candidate, major research interests include the application of modern control theory in power electronics and distributed generation. Li Huiguang (1947-), male, from Qiqihar City, Professor, PhD supervisor, major research interests include sampling theory, robot vision, distributed generation, and renewable energy.