With the extensive applications of biomagnetic signals derived from active biological tissue in both clinical diagnoses and human-computer-interaction, there is an increasing need for approachable weak biomagnetic sensing technology. The inherent merits of giant magnetoresistance (GMR) and its high integration with multiple technologies makes it possible to detect weak biomagnetic signals with micron-sized, non-cooled and low-cost sensors, considering that the magnetic field intensity attenuates rapidly with distance. This paper focuses on the state-of-art in integrated GMR technology for approachable biomagnetic sensing from the perspective of discipline fusion between them. The progress in integrated GMR to overcome the challenges in weak biomagnetic signal detection towards high resolution portable applications is addressed. The various strategies for 1/f noise reduction and sensitivity enhancement in integrated GMR technology for sub-pT biomagnetic signal recording are discussed. In this paper, we review the developments of integrated GMR technology for in vivo/vitro biomagnetic source imaging and demonstrate how integrated GMR can be utilized for biomagnetic field detection. Since the field sensitivity of integrated GMR technology is being pushed to fT/Hz0.5 with the focused efforts, it is believed that the potential of integrated GMR technology will make it preferred choice in weak biomagnetic signal detection in the future.

Accurate prediction of binding affinity between protein and ligand is a very important step in the field of drug discovery. Although there are many methods based on different assumptions and rules do exist, prediction performance of protein–ligand binding affinity is not satisfactory so far. This paper proposes a new cascade graph-based convolutional neural network architecture by dealing with non-Euclidean irregular data. We represent the molecule as a graph, and use a simple linear transformation to deal with the sparsity problem of the one-hot encoding of original data. The first stage adopts ARMA graph convolutional neural network to learn the characteristics of atomic space in the protein–ligand complex. In the second stage, one variant of the MPNN graph convolutional neural network is introduced with chemical bond information and interactive atomic features. Finally, the architecture passes through the global add pool and the fully connected layer, and outputs a constant value as the predicted binding affinity. Experiments on the PDBbind v2016 data set showed that our method is better than most of the current methods. Our method is also comparable to the state-of-the-art method on the data set, and is more intuitive and simple.

As a preliminary study for bearingless permanent magnet slice motor (BPMSM) development, an effective means for BPMSM mechanical structure optimization is proposed here by developing a virtual prototype based on Ansoft Maxwell to realize overall performance improvements. First, the sensitivity evaluation index of the candidate mechanical structural parameters for individual BPMSM performance is constructed for selection. Orthogonal tests are performed to determine the dominant mechanical structural parameters to be optimized by utilizing monitored data based on Ansoft Maxwell. A linear regression model of the mechanical structural parameters for specific performances is obtained by utilizing the gradient descent method. Then, a multi-structural optimization regression model of the selected dominant mechanical structural parameters for overall performance is established using an analytic hierarchy process and solved using a genetic algorithm. The simulation results show that the performance of the optimized BPMSM has been comprehensively improved. Specifically, the passive axial stiffness, passive tilting stiffness, force-current coefficient, and motor efficiency increased by 56.4%, 71.3%, 19.6%, and 8.7%, respectively.

No abstract

No abstract