云過(guò)程神經(jīng)網(wǎng)絡(luò)模型及算法研究
doi: 10.11999/JEIT140329 cstr: 32379.14.JEIT140329
基金項(xiàng)目:
國(guó)家自然科學(xué)基金(61170132)和中國(guó)博士后科學(xué)基金(201003405)資助課題
Research on Cloud Process Neural Network Model and Algorithm
-
摘要: 該文針對(duì)輸入輸出具有不確定性特征并與時(shí)間或過(guò)程有關(guān)的復(fù)雜非線性系統(tǒng)建模和求解問(wèn)題，利用過(guò)程神經(jīng)網(wǎng)絡(luò)對(duì)時(shí)變信號(hào)的動(dòng)態(tài)處理能力，結(jié)合云模型對(duì)定性定量概念的轉(zhuǎn)化能力，構(gòu)建了一種具有不確定性信息處理能力的云過(guò)程神經(jīng)網(wǎng)絡(luò)模型，并采用貓群優(yōu)化算法同時(shí)對(duì)網(wǎng)絡(luò)結(jié)構(gòu)和參數(shù)進(jìn)行并行優(yōu)化設(shè)計(jì)，提高了網(wǎng)絡(luò)逼近及泛化能力，實(shí)現(xiàn)了神經(jīng)網(wǎng)絡(luò)在時(shí)間域和不確定信息處理領(lǐng)域上的有效擴(kuò)展。仿真實(shí)驗(yàn)結(jié)果驗(yàn)證了模型和算法的可行性和有效性。
-
關(guān)鍵詞:
- 云模型 /
- 云過(guò)程神經(jīng)網(wǎng)絡(luò) /
- 貓群優(yōu)化 /
- 時(shí)間序列預(yù)測(cè)
Abstract: For modeling and solving problems of complex nonlinear systems whose input/output have uncertainty and are associated with time or process, a cloud process neural network model is built in the paper. It has uncertainty information processing ability by combining process neural networks processing ability for time-varying signal with cloud model transformation ability between qualitative and quantitative concepts. In addition, the cat swarm optimization algorithm is used to optimize the network structure and parameters simultaneously, and it helps to improve network approximation?and generalization performance. The effective extension of neural networks in time domain and uncertain information processing field is realized. Experimental results verify the effectiveness and feasibility of the model and algorithm. -
計(jì)量
- 文章訪問(wèn)數(shù): 1917
- HTML全文瀏覽量: 116
- PDF下載量: 664
- 被引次數(shù): 0
下載:
