基于神經(jīng)網(wǎng)絡(luò)的太陽(yáng)電池陣熱真空試驗(yàn)外熱流模擬系統(tǒng)辨識(shí)
The solar battery array thermal vacuum test model identification based on neural network
-
摘要: 文章針對(duì)太陽(yáng)電池陣熱真空試驗(yàn)非線性的特點(diǎn)，引入一種基于神經(jīng)網(wǎng)絡(luò)的系統(tǒng)辨識(shí)方法。該方法采用BP網(wǎng)絡(luò)作為模型辨識(shí)器，而辨識(shí)器又采用L-M（Levenberg-Marquart）算法進(jìn)行訓(xùn)練。仿真結(jié)果表明，該方法具有較高的訓(xùn)練速度與精度，可以對(duì)太陽(yáng)電池陣熱真空試驗(yàn)測(cè)點(diǎn)溫度響應(yīng)做出較為精確的預(yù)測(cè)。Abstract: In view of the nonlinear characteristics of the solar cell array in thermal vacuum tests, an identification method based on the neural network is proposed. The neural network is trained by the Levenberg-Marquart algorithm. The simulation results show that the system can achieve a fast identification and with a high precision, and the temperature response at the measurement point on the solar array can be predicted.
下載: