基于Kalman濾波的Helmholtz機(jī)進(jìn)化算法
Helmholtz Machine Evolutionary Algorithm Based on Kalman Filter
-
摘要: 根據(jù)信息融合的思想，提出了基于Kalman濾波的Helmholtz進(jìn)化算法(KFHMEA)。在KFHMEA中，首先將種群分成若干子種群，各子種群獨(dú)立使用Helmholtz機(jī)來(lái)估計(jì)其狀態(tài)，然后用Kalman濾波器把這些狀態(tài)信息相融合，并將融合后的信息反饋到各子種群中。仿真實(shí)驗(yàn)表明，這種Kalman融合機(jī)制能夠大大增強(qiáng)原有算法的穩(wěn)定性和全局搜索能力，從而提高了算法的整體性能。Abstract: On the basis of information fusion, an Evolutionary Algorithm based on Helrnholtz Machine and Kalman Filter (KFHMEA) is proposed. In KFHMEA, the population is divided into several sub-populations and Helmholtz machine is used to estimate the state of each sub-population. Then the information of these states is fused using a Kalman filter, and the fused information is feed back to the sub-populations. Experimental results demonstrate that KFHMEA outperforms HMEA in the stability and global search ability.
-
計(jì)量
- 文章訪問(wèn)數(shù): 2222
- HTML全文瀏覽量: 134
- PDF下載量: 428
- 被引次數(shù): 0
下載:
