| According to the requirements of "the 14 th Five Year Plans for National Economic and Social Development and the Long-Range Objectives Through the Year 2035",the new energy storage is an important technology for building a new power system,and an important support for building a clean,low-carbon,safe and efficient energy system and achieving the goals of carbon peaking and carbon neutrality.As a new type of energy storage equipment,the flywheel energy storage systems have the advantages of high energy storage density,large instantaneous power,fast charging and discharging,long service life,high energy conversion efficiency,and are widely used in new energy vehicles,distributed generation,uninterruptible power supply and other fields.Magnetic bearings are widely used in flywheel energy storage support systems to solve the problems of friction,wear,and lubrication of mechanical bearings.The use of the magnetic bearing,however,lengthens the original system’s axial length and restricts the flywheel’s critical speed and output power.The bearingless permanent magnet synchronous motor,a new kind of motor that combines magnetic bearing and permanent magnet synchronous motor,has features like a compact structure,lightweight,high critical speed,high output power,etc.,that make it ideal for flywheel energy storage support systems.The outer rotor coreless bearingless permanent magnet synchronous motor(ORC-BPMSM)adopts an outer rotor and coreless structure on the basis of the bearingless permanent magnet synchronous motor,which has the advantages of large rotational inertia,low suspension power consumption,and low core loss,which is more conducive to improving the comprehensive performance of the flywheel energy storage devices.The researches presented in this dissertation have significant implications for advancing flywheel energy storage’s fundamental technology,removing the technology’s bottleneck,and enhancing China’s flywheel energy storage industry’s engineering application capabilities and global competitiveness.The ORC-BPMSM is used as an investigation object in the dissertation,and research is conducted on its operating concepts,structural parameters optimization design,mathematical models,parameter identification,active disturbance rejection decoupling control methods,unbalanced vibration of the rotor,and digital control systems.The primary contributions and accomplishments of the dissertation are as follows:1.A surrogate model based on radial basis function neural network and a multi-objective robust optimization algorithm are proposed to design the initial parameters of stator and rotor in the ORC-BPMSM.The operation of the ORC-BPMSM is explained using an analysis of the electromagnetic force existing in the system.On the basis of analyzing the electromagnetic force of ORC-BPMSM,the working principle of ORC-BPMSM is elaborated and its mathematical model is established.The response surface model is used to establish the functional relationship between the optimization objectives and optimization variables,and the optimal solution set is obtained through the deterministic optimization algorithm.The response surface model is sampled by the Monte Carlo method,and a robust optimization algorithm is proposed to obtain a new optimal solution set.The failure probabilities of deterministic optimization and robust optimization are compared and analyzed to verify the correctness and effectiveness of the proposed method.The experimental results show that the proposed multi-objective robust optimization algorithm can effectively improve the torque and suspension performance of the ORC-BPMSM.2.A linear active disturbance rejection decoupling control method based on adaptive neural network parameter identification algorithm is proposed to address the coupling problem between torque and suspension force in the ORC-BPMSM control process.According to the flux linkage and voltage equations of the motor side and the suspension force side in the ORC-BPMSM,the parameters to be identified are analyzed.The static identification method is used for the slowly changing permanent magnet flux linkage parameters and mutual inductance coefficients,and the adaptive neural network algorithm is used for real-time adjustment of the rapidly changing motor side inductance parameters and suspension force side inductance parameters.After that,the identified permanent magnet magnetic linkage parameters,mutual inductance coefficient,motor side inductance parameters,and suspension force side inductance parameters are applied to subsequent control algorithms.According to the ORC-BPMSM mathematical model,the linear active disturbance rejection controller(LADRC)is designed for the inner and outer loops of the motor side and the suspension force side respectively.Finally,the proposed algorithm is validated on an experimental platform.The experimental results show that the proposed parameter identification LADRC decoupling control method can effectively eliminate the coupling between torque and suspension force.3.A genetic algorithm(GA)and the back-propagation neural network(BPNN)algorithm are proposed to address the issue of fixed coefficients affecting control performance in LADRC decoupling control methods.This method uses the GA based on real number encoding to optimize the initial value of the BPNN,avoiding the problem of system runaway caused by convergence failure of the BPNN.Meanwhile,this method has the advantage of smaller computational complexity and can meet real-time requirements.Finally,the proposed algorithm is validated on the experimental platform.The experimental results show that the LADRC parameter dynamic adjustment algorithm not only improve the control performance of the control system but also reduce the failure probability of the control system during the parameter dynamic adjustment process.4.An adaptive neural network band-pass filter vibration compensation algorithm is proposed to address the rotor vibration problems during the operation of the ORC-BPMSM.The mechanism of rotor unbalance vibration and dead time vibration in the ORC-BPMSM is analyzed.The adaptive neural network bandpass filter algorithm is used to extract the harmonic components of unbalanced vibration and dead time vibration,respectively.The extracted harmonic components are adjusted by a PID controller to tend toward zero.Afterward,the adjusted compensation amount is injected into the forward channel to complete the compensation control.The proposed vibration compensation algorithm is verified on the experimental platform.The experimental results show that the proposed vibration compensation method not only effectively suppresses unbalanced vibration,but also reduces the harmonic amplitude of dead time vibration,and the performance of the control system is greatly improved.5.The optimized ORC-BPMSM prototype is manufactured,and the ORC-BPMSM control system with TMS320F28335 as the control core is designed.Based on the parameter identification,the LADRC method,the GA-BPNN parameter adjustment algorithm,and the vibration compensation algorithm proposed in this dissertation,not only the comprehensive experimental verification of the algorithm is conducted on the experimental platform,but also the fundamental testing of the ORC-BPMSM power generation performance is conducted,which provides a theoretical basis and experimental basis for the application of the ORC-BPMSM in the corresponding field.The experimental results show that the integrated control algorithm has excellent vibration suppression ability and good operation effect,and the fluctuation of the direct current bus voltage and displacement can meet the control requirements in the power generation state. |
PDF Full Text Request