Emotional Option Strategy Research

Social emotional optimization algorithm (SEOA) is a recently proposedswarm intelligent algorithm by simulating the decision process among humansociety. In SEOA, each individual denotes one virtual person, and selects thebehavior motion according to emotional index. Emotional index is one crucialparameter with linear updating manner. In this paper, the selection strategy ofemotional index is discussed, and the main contributions are listed as follows:(1) In real world, the emotion of one person is always influenced by manyrandom elements; therefore, a stochastic emotion is more fit for the real personthan a determined one. However, the stochastic character is omitted in thestandard version of SEOA. By the way, up to now, much phenomenon amonganimal group is determined with Levy distribution. Therefore, we incorporateLevy distribution into the selection strategy of emotional index, and call it asSEOA based on Levy distribution (SEOA-LD), simulation results show it isbetter than the standard version.(2) With large number law, the effect of amount random events is the sameas one normal distribution. So, inspired by the Levy distribution, we consider tointroduce normal distribution into the selection strategy of emotional index. Inthis strategy, each individual owns one normal distribution with mean value0,and adjusts its standard according to different behaviors. However, due tothe3principle for normal distribution, two extreme emotions is occurred in aspecial low probability (less than0.03%). Furthermore, the shape of Cauchydistribution is similar with normal distribution; however, the trait of Cauchydistribution is with a larger probability than that of normal distribution.Therefore, we incorporate normal distribution and Cauchy distribution intoSEOA, respectively. Simulation results show SEOA with normal distribution(SEOA-ND) is superior to other two random selection strategies.(3) To testify the performance of SEOA-ND, we apply it to solve thedirecting orbits of chaotic systems, and compare with standard particle swarmoptimization and standard version of SEOA. Simulation results show SEOA-NDis superior to other two algorithms.