Design of Highly Uniform Magnetic Field Coils With Wolf Pack Algorithm

Jun Xu,Xiangyu Kang,Zhengkun Fan,Zhanfei Zhang,Yang Li,Xiangmei Dong,Xiumin Gao

DOI: https://doi.org/10.1109/jsen.2020.3030622

IF: 4.3

2021-02-15

IEEE Sensors Journal

Abstract:Uniform magnetic fields are widely used in atomic magnetometers, MRI and other fields. The traditional method of solving closed-form equations is usually used to design highly uniform magnetic field coils. In this paper, a new intelligent algorithm, wolf pack algorithm (WPA), is proposed to design the coil structure of a highly uniform magnetic field. This method can replace the traditional method to obtain the coil parameters. We additionally use a discrete optimization method to solve the rounding error. According to the investigation, compared with particle swarm optimization (PSO) algorithm, WPA has stronger global optimization ability, faster convergence speed and multiple optimization strategies. We set appropriate coil constraints (structural and process constraints) and simplify the multi-dimensional solutions to one-dimensional solutions to reduce the difficulty of operation. The distribution of magnetic field intensity of three axes of coils is analyzed by the finite element analysis method, and the 1%, 0.1%, and 0.01% uniform regions are studied. Under the same conditions, the uniform magnetic field regions of the coils designed by the PSO and WPA, as well as a single pair of axial and radial coils are compared. The performance results of multiple pairs of coils are superior to those of a single pair of coils. In general, the uniformity of the axial and radial magnetic field coils designed by the WPA is better than that designed by the PSO. Finally, the experimental value of the magnetic field uniformity is nearly consistent with the theoretical value.

engineering, electrical & electronic,instruments & instrumentation,physics, applied

What problem does this paper attempt to address?

This paper aims to solve the problem of how to design a coil structure that generates a highly uniform magnetic field. Specifically, the paper proposes a new intelligent algorithm - the Wolf Pack Algorithm (WPA) - for designing a coil structure capable of generating a highly uniform magnetic field. Traditional methods usually use analytical equations to design magnetic field coils with high uniformity, while the WPA proposed in this paper can replace traditional methods to obtain coil parameters and performs excellently in terms of global optimization ability, convergence speed, and multiple optimization strategies. In addition, in order to reduce rounding errors, a discrete optimization method is also used in the research.

By setting appropriate coil constraints (such as structural and technological constraints), the paper simplifies multi - dimensional solutions into one - dimensional solutions to reduce the difficulty of operation. The magnetic field intensity distributions of three axial coils are analyzed by the finite - element analysis method, and the uniform regions of 1%, 0.1% and 0.01% are studied. Under the same conditions, the uniform magnetic field regions of coils designed by the Particle Swarm Optimization (PSO) algorithm and WPA, as well as single - pair axial and radial coils, are compared. The results show that the performance of multi - pair coils is better than that of single - pair coils, and the uniformity of axial and radial magnetic field coils designed by WPA is better than that of coils designed by PSO. Finally, the experimental results are basically consistent with the theoretical values.

In summary, this paper mainly solves the problem of how to use WPA to efficiently and accurately design a coil structure that can generate a highly uniform magnetic field, which is of great significance for improving the accuracy of devices such as atomic magnetometers.