ass日本风韵熟妇pics男人扒开女人屁屁桶到爽|扒开胸露出奶头亲吻视频|邻居少妇的诱惑|人人妻在线播放|日日摸夜夜摸狠狠摸婷婷|制服 丝袜 人妻|激情熟妇中文字幕|看黄色欧美特一级|日本av人妻系列|高潮对白av,丰满岳妇乱熟妇之荡,日本丰满熟妇乱又伦,日韩欧美一区二区三区在线

As the environmental factors were only considered in the existing photosynthetic rate prediction models based on neural network, slow convergence speed was still the existing problem. The temperature, CO2 concentration, photon flux density and relative humidity, especially the chlorophyll content were considered. Photosynthetic rate prediction model of cucumber seedlings fused chlorophyll content was proposed. Firstly, 825 experimental data of cucumber seedlings photosynthetic rate were obtained by multi-factor coupling test. The temperature gradients were set at 16, 20, 24, 28, 32℃, respectively, CO2 concentration gradients were set at 300, 600, 900, 1 200, 1 500 μL/L and the photon flux density gradients were set at 0, 20, 50, 100, 200, 300, 500, 700, 1 000, 1 200, 1 500 μmol/(m2·s), respectively. Secondly, Levenberg-Marquardt (LM) training method was used. Meanwhile, the effect of chlorophyll content on the training results was analyzed. Then different calibrations were used to validate the multi-factor coupling photosynthetic rate prediction model. The results showed that the training results of the training method considered chlorophyll content and the model fitting degree were superior to the training model only considered the environmental factors. Because of the local area, LM training method considering chlorophyll content can effectively flat over the local area and meet the training requirement. The error rate was less than 0.000 1 and the determination coefficient between actual measured and calculated values was 0.987. It indicated that these two values had good correlation and similarity. Besides, the error was less than 4.68%, which proved that the proposed model has a high accuracy.