Determination of Net Photosynthetic Rate of Plants Based on
Environmental Compensation Model
LI Cong-cong1, LUO Qi-wu2, ZHANG Ying-ying1, 3*
1. School of Electrical Engineering and Automation, Hefei University of Technology, Hefei 230009, China
2. School of Automation, Central South University, Changsha 410083, China
3. National and Local Joint Engineering Laboratory of Renewable Energy Access to Power Grid Technology (Hefei University of Technology), Hefei 230009, China
Abstract：Increasing the photosynthetic rate of crops is one of the reliable methods for high yield breeding. The main method for measuring photosynthesis rate is infrared gas analysis, which owns dependable axiom and mature technology. However, the infrared light source is easily affected by the complex working environment in the field, especially the change in ambient temperature. Therefore, the measurement error is significant in the task of quantitative analysis, and the detection precision of gas with deficient concentration or weak concentration change is not exact. Based on the above questions, first of all, the tunable diode laser absorption spectroscopy (TDLAS) is applied to the measurement of plant photosynthetic rate in this paper, which employs the second harmonic peak difference to represent the relative variation of trace concentration of photosynthetic gas CO2 in unit sampling time. Secondly, we established an environment compensation model of a broad learning system based on firefly algorithm optimization (FA-BLS). The position information of each firefly in the model corresponds to a set of feasible solutions representing the weights and thresholds of the BLS. Through the continuous iteration and update optimization of firefly position to find the firefly with the highest brightness, that is to generate the weights and thresholds that make the model perform the best. Ultimately, the compensation value generated by FA-BLS is used to compensate for the original second harmonic peak difference with environmental impact, and the net photosynthetic rate per unit sampling time was obtained from the compensated second harmonic peak difference. The experimental results indicate that the firefly population size and the number of nodes in the enhancement layer of BLS are significant considerations affecting the output error of TDLAS-FA-BLS, which commendably inherits the advantages of BLS, such as fast training speed and short iteration time. It is worth mentioning that the average measurement time of FA-BLS is merely 0.81 s, and the chi square distance between model prediction output and test set data is only 0.29×10-4, which indicates its output error is similarly small. At the same time, the sample variance and sample standard deviation of the output error of FA-BLS are lower than those of BLS, which illustrates that FA-BLS overcomes the shortcomings of BLS, such as unstable network output and low generalization due to random selection of parameters. Consequently, the method based on TDLAS-FA-BLS for the determination of plant net photosynthetic rate can nicely meet the needs of high precision, real-time, stability and reliability in the complex field working environment and actual agricultural production.
李聪聪，罗旗舞，张莹莹. 基于环境补偿模型的植物净光合速率测定[J]. 光谱学与光谱分析, 2022, 42(05): 1561-1566.
LI Cong-cong, LUO Qi-wu, ZHANG Ying-ying. Determination of Net Photosynthetic Rate of Plants Based on
Environmental Compensation Model. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2022, 42(05): 1561-1566.
[1] Orr D J, Alcantanra A, Kapralov M V, et al. Plant Physiology, 2016, 172(2): 707.
[2] Yin J, Liu X Y, Miao Y L, et al. International Journal of Agricultural and Biological Engineering, 2019, 12(5): 156.
[3] Yin G F, Zhao N J, Shi C Y, et al. Optics Express, 2018, 26(6): A293.
[4] Du T T, Meng P, Huang J L, et al. Plant Methods, 2020, 16(1): 10.
[5] LIU Zi-huai, YANG Chun-hua, LUO Qi-wu, et al（刘紫怀，阳春华，罗旗舞，等）. Acta Optica Sinica（光学学报）, 2019, 39(10): 335.
[6] JU Yu, CHEN Hao, HAN Li, et al（鞠 昱，陈 昊，韩 立，等）. Spectroscopy and Spectral Analysis（光谱学与光谱分析）, 2020, 40(12): 3665.
[7] Zhang Z R, Pang T, Yang Y, et al. Optics Express, 2016, 24(10): 13.
[8] Xin M Y, Song J L, Rao W, et al. Chinese Journal of Aeronautics, 2020, 33(12): 3158.
[9] Chen C L P, Liu Z L. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(1): 10.
[10] Zhao G X, Wang X S, Kong Y, et al. Remote Sensing, 2021, 13(4): 23.
[11] Han J, Xie L, Liu J, et al. Multimedia Tools and Applications, 2020, 79(23-24): 16627.
[12] LI Yi-kai, ZHANG Tong, CHEN C L Philip（李逸楷，张 通，陈俊龙）. Acta Automatica Sinica（自动化学报）, 2020, 46(10): 2060.
[13] Sheng B, Li P, Zhang Y H, et al. IEEE Transactions on Cybernetics, 2021, 51(3): 1463.
[14] Yu W K, Zhao C H. IEEE Transactions on Industrial Electronics, 2020, 67(6): 5081.
[15] Yang X S. Engineering with Computers, 2013, 29(2): 175.
