【摘要】：針對(duì)航空肼燃料保障安全評(píng)價(jià)的復(fù)雜性和非線性,提出并建立了基于BP和Hopfield神經(jīng)網(wǎng)絡(luò)的動(dòng)態(tài)安全評(píng)價(jià)模型。在綜合分析國內(nèi)外肼燃料保障安全評(píng)價(jià)的基礎(chǔ)上,針對(duì)航空肼燃料保障過程中出現(xiàn)的問題,構(gòu)建并優(yōu)化了指標(biāo)體系,選取前饋神經(jīng)網(wǎng)絡(luò)中的BP網(wǎng)絡(luò)和反饋神經(jīng)網(wǎng)絡(luò)中的Hopfield網(wǎng)絡(luò)建立評(píng)價(jià)模型。在詳細(xì)說明了BP和Hopfield神經(jīng)網(wǎng)絡(luò)的構(gòu)建方法后,進(jìn)行實(shí)例驗(yàn)證,并對(duì)預(yù)測(cè)效果進(jìn)行了比較分析。仿真表明,兩種模型都能正確評(píng)價(jià)安全保障狀態(tài)。但在收斂速度、聯(lián)想記憶功能方面Hopfield神經(jīng)網(wǎng)絡(luò)優(yōu)于BP神經(jīng)網(wǎng)絡(luò)。將BP和Hopfield神經(jīng)網(wǎng)絡(luò)用于肼燃料保障安全評(píng)價(jià)過程中,具有適用性和可行性,對(duì)于航空肼燃料保障的安全建設(shè)與安全管理研究具有重要意義。
[Abstract]:In view of the complexity and nonlinearity of aviation hydrazine fuel safety evaluation, a dynamic safety evaluation model based on BP and Hopfield neural network is proposed and established. On the basis of synthetically analyzing the safety evaluation of hydrazine fuel support at home and abroad, and aiming at the problems in the process of aviation hydrazine fuel support, the index system is constructed and optimized. BP network in feedforward neural network and Hopfield network in feedback neural network are selected to establish evaluation model. The construction method of BP and Hopfield neural networks is explained in detail, and the results of prediction are compared and analyzed. Simulation results show that both models can correctly evaluate the security state. But Hopfield neural network is superior to BP neural network in convergence speed and associative memory function. The application of BP and Hopfield neural networks to the safety evaluation of hydrazine fuel support is applicable and feasible, which is of great significance to the safety construction and safety management of aviation hydrazine fuel support.
【作者單位】： 空軍勤務(wù)學(xué)院;
【分類號(hào)】：TP183;V312
，

本文編號(hào)：2144873