This paper describes an artificial neural net based approach to prediction of power system harmonic voltages. Recurrent neural networks are used to handle harmonics dynamics. A comparison is made of four recurrent network models from a standpoint of the accuracy and computational efforts.

Objective: To assess the effectiveness of color images as mood stimulants by using an image selection system. Methods: Color images of flowers and seascapes, each mainly comprising of 5 colors, were arranged into a set of 50 images (10 images for each of 5 colors) on a personal computer and shown to 40 university students on stressful days during (i) an examination period and (ii) a non-examination period. The students selected color images according to their individual preferences by using an image selection system. To determine the effect of viewing the color images, mood adjective scores as determined by the Multiple Mood Scale (MMS) were measured as an emotional parameter, and the levels of salivary chromogranin A (CgA) and salivary cortisol were measured as biochemical parameters. These data were analyzed statistically and a p value less than 0.05 was considered significant. Results: After viewing the color images, mean MMS scores indicating positive moods, such as well-being and friendliness increased significantly (p<0.05), and scores indicating negative moods, such as depression, boredom and hostility decreased significantly(p<0.05). The mean value of salivary CgA also decreased significantly after viewing the images (p<0.05). Conclusion: Color images could possibly be used to induce a parasympathetically dominant state.

No abstract