Electromagnetic fields (EMFs) have attracted increasing concerns with the exponential development of electrical appliances, such as electric vehicles (EV), 5G technology, the Internet of Things, etc. International organizations such as ICNIRP and IEEE have already stringently reviewed and published general exposure limits related to EMFs. However, research on the impact of electromagnetic radiation on human health and safety is ongoing, contributing to the protection of human safety under electromagnetic exposure.

Dosimetry studies, which evaluate the magnitude and distribution of in situ EMFs inside the human body, provide scientific evidence for precise derivation and supplement the general exposure limits in the EMF standards. This section first reviews the changes and reasoning of the ICNIRP and IEEE electromagnetic radiation safety standards for the human body, examines the human models and computational techniques used in dosimetry studies, and gives an example calculation.

The effects of static magnetic fields (SMF) produced by EVs on human neuro-psychological responses and brain activity are analyzed. It is widely believed that abnormalities in human neuro-psychological responses and brain activity during driving are causes of traffic and public safety problems, and SMFs may stimulate these. A lane change task is examined to evaluate driving performance. A driving reaction time test and a reaction time test are conducted to assess the variation of the neuro-psychological cognitive functions. Both sham and actual exposure conditions are examined with a localized SMF from an EV; student subjects are considered in this single-blind experiment. Electroencephalographs (EEG) of the subjects are recorded during the experiment as an indicator of brain activity for variations in driving performance and cognitive functions. Results of this study indicate that the impact of the given SMF on both human driving performance and cognitive functions is inconsiderable. There is a correlation between the beta sub-band of the EEGs and the human reaction time.

A new hybrid forecasting model for periodic time-series signals is introduced. Human physiological and electrophysiological signals, such as EEGs, electrocardiography, blood glucose concentration, etc., can be considered as non-linear and non-stationary time series with periodic characteristics. Due to experimental conditions, it is not easy to obtain reliable human electrophysiological signals under different electromagnetic exposure conditions. Therefore, reliable periodic power consumption data are selected in this study to verify the effectiveness of the hybrid forecasting model. Based on the analysis and 'decompose and ensemble' strategy, a hybrid network named 'iCEEMDAN-BOA-GRU' based on the improved complete ensemble empirical mode decomposition with adaptive noise (iCEEMDAN), Bayesian optimization algorithm (BOA), and gated recurrent unit (GRU) is proposed to achieve accurate short-term load prediction of urban functional zones (UFZ). First, the raw power consumption data is decomposed into a series of patterns with noticeable differences through iCEEMDAN. Then, a Bayesian-optimized GRU is used to predict each mode individually. Finally, the prediction results of each mode are superimposed and reconstructed to form an overall prediction result. In each training, the BOA is used to select the best-fitting GRU hyperparameters to match the data characteristics of each model. An actual UFZ power consumption dataset is adopted to perform the simulation. The results show that the proposed new hybrid model can accurately predict the power consumption data of UFZs and offers the highest prediction ability of all survey models compared to the prediction errors of other models. The proposed hybrid prediction model also has excellent potential to be extended to an effective tool for accurately predicting human biological signals under different electromagnetic exposure conditions.

A novel model based on the support vector machine with a radial basis function kernel (RBF-SVM) using time-series features of zebrafish (Danio rerio) locomotion exposed to different EMFs is introduced. It is used to indicate the corresponding EMF exposure. A group of zebrafish is randomly and evenly divided into two groups; each group of fish is subjected to a new tank test at 6.78 MHz and sham or real magnetic exposures of approximately 1 A/m. Their videotaped locomotion in the tests is converted into x and y time-series coordinates of the trajectories, then reformed into time-series matrices according to different time-series lengths. Features of zebrafish locomotion are calculated using the highly comparative time-series analysis (HCTSA). A limited number of time-series features most relevant to the EMF exposure conditions are selected using the minimum-redundancy maximum-relevance algorithm for RBF-SVM classification training. It has previously been quantitatively verified that ambient environmental parameters have little effect on the locomotion performance of zebrafish processed by the empirical method. The results demonstrate that the proposed model is capable of accurately indicating different EMF exposures. All classification accuracies are 100% and the classification precisions of several classifiers based on specific parameters and feature sets with specific dimensions can reach higher than 95%. The reason for this result is believed to be that the specified EMF has affected the zebrafish's neural features, which is reflected in their behaviors. The outcomes of this study provide a new indication model for EMF exposures and provide a reference for investigation of the impact of EMF exposure.

A new rating model for measurement of the dielectric properties of human tissues under different conditions is proposed. Using the dielectric properties of human tissues, the results of dosimetry studies are used to evaluate the intensity and distribution of in situ electromagnetic fields in the human body, providing scientific evidence for the accurate derivation and supplementing the general exposure limits in the EMF standards. Therefore, selecting electrical parameters appropriate for the human body is essential for assessing human exposure to electromagnetic radiation. Based on the current measurement results of the electrophysiological parameters of human tissues, this study establishes a new rating model for measurement of the dielectric properties of human tissues under different conditions. The purpose of establishing this model is to better support accurate estimation of the effects of electromagnetic radiation on the human body and provide a reference for establishing the new standard.