臺灣博碩士論文加值系統

根據現今植物感測研究不斷受到重視，目的不是為了簡單地表明「植物有思維」，而是在於探討植物為什麼以及如何感知周圍的環境。動植物一直以來均會使用富含特定化學物質的氣味與外界互相交流的訊號，其中植物會透過化學物質來進行交流、生長、防禦的例子的課題。
植物電生理傳感器，可呈現植物思維的自主設備，可以傳送植物潛在數據轉化為植物自主意識訊號。植物傳感可以測量某些方面天然植物內部狀態，傳感器電路設計於心電圖和腦電圖測量電生理學的方法，本論文說明提出藉由使用植物電生理方法來收集生理數據，對植物本身的行為做歸納數據分析，介紹環境與動作因素對植物產生的內部生理電反饋訊號。本文以其眾多神經網路使用長短期記憶【 Long Short Term Memory Network,LSTM】神經網路分析，因其LSTM 神經元在其中保留權重數據上下文，以允許解決植物生理反應和時間序列問題，而不會出現梯度消失問題影響，解決生理時序連續變化關聯性，可分析出當前動作對複雜的連續生理訊號關係。

According to the current research on plant sensing has been receiving more and more attention, the purpose is not to simply show that "plants have thought", but to explore why and how plants perceive the surrounding environment. Animals and plants have always used odors rich in specific chemical substances to communicate with the outside world. Among them, plants communicate, grow, and defend through chemical substances.
Plant electrophysiological sensors are autonomous devices that can present plant thinking, and can transmit plant potential data into plant autonomous consciousness signals. Plant sensing can measure some aspects of the internal state of natural plants. The sensor circuit is designed to measure electrophysiology in electrocardiography and electroencephalography. This paper demonstrates that by using plant electrophysiology methods to collect physiological data, the behavior of plants themselves Do inductive data analysis and introduce the internal physiological electrical feedback signals of plants produced by environmental and action factors. This paper uses Long Short Term Memory [Long Short Term Memory Network, LSTM] neural network analysis with its many neural networks, because its LSTM neurons retain weight data context in it to allow solving plant physiological responses and time series problems without There will be the influence of the gradient disappearance problem, solve the correlation of continuous changes in physiological time series, and analyze the relationship between current actions and complex continuous physiological signals.

摘要..................................i
ABSTRACT...............................ii
誌謝...............................iii
目錄...............................iv
表目錄...............................vi
圖目錄...............................vii
第一章、序論...............................1
1.1、研究背景...............................1
1.2、研究動機...............................2
1.3、研究方法...............................3
1.4、論文組織...............................4
第二章 背景與相關知識...............................5
2.1 植物電生理解析...............................5
2.1.1、植物離子通道...............................6
2.1.2、動作電位 (Action Potential)...............................7
2.1.3、變異勢 Variation Potential...............................8
2.1.4、植物週期性生理電節奏...............................9
2.2 前饋神經網絡 (FFNN)...............................10
2.3 、RNN(RECURRENCE NEURAL NETWORK)...............................11
2.3.1、RNN 隱藏層傳遞狀態...............................12
2.3.2、輸入權重衝突 （input weight conflict）...............................13
2.3.3、反向傳播（Backpropagation）...............................13
2.4、RNN的SEQ2SEQ模型...............................14
2.4.1、時間卷積網絡(Temporal Convolution Network)..............................15
第三章 系統設計與研究方法...............................16
前言...............................16
3.1、植物傳感模組...............................17
3.2、電路板量測原理...............................18
3.2.1、準同步採樣原理...............................18
3.2.2、因模組轉換器...............................20
3.3、探針量測考量與使用...............................21
3.3.1、針電極...............................21
3.4、實驗數據測試...............................22
3.5、LSTM神經網絡分析...............................23
3.5.1、數據收集方法...............................24
3.6.1、原始數據特徵值...............................25
3.6.2、數據集參數...............................26
3.6.2、LSTM 矩陣運算...............................28
3.6.3、向量維數...............................29
第四章 實驗分析與結果...............................30
前言...............................30
4.1 、環境架設..............................31
4.1.1植物生長燈..............................32
4.2 植物電生理資料集..............................33
4.2.1、植物數據集觀察..............................34
4.2.2、資料整理..............................36
4.3、特徵工程..............................37
4.3.1 、One-hot編碼..............................38
4.3.2、雜訊處理..............................38
4.3.3、移動標準差提取特徵..............................39
4.3.4、特徵連續變量..............................40
4.4、LSTM訓練模型..............................41
第五章 結論..............................42
第六章 未來展望..............................43
參考文獻..............................44

[1]H. Herzog, B. Eliasson, and O. Kaarstad, "Capturing greenhouse gases," Scientific American, vol. 282, no. 2, pp. 72-79, 2000.
[2]B. Cooke, "Disease assessment and yield loss," in The epidemiology of plant diseases: Springer, 2006, pp. 43-80.
[3]C. J. Love, S. Zhang, and A. Mershin, "Source of sustained voltage difference between the xylem of a potted Ficus benjamina tree and its soil," PloS one, vol. 3, no. 8, p. e2963, 2008.
[4]S. K. Chatterjee et al., "Exploring strategies for classification of external stimuli using statistical features of the plant electrical response," Journal of the Royal Society Interface, vol. 12, no. 104, p. 20141225, 2015.
[5]罗海波, 许凌云, 惠斌, and 常铮, "基于深度学习的目标跟踪方法研究现状与展望," 红外与激光工程, vol. 46, no. 5, p. 502002, 2017.
[6]F. Bezanilla, "The voltage sensor in voltage-dependent ion channels," Physiological reviews, vol. 80, no. 2, pp. 555-592, 2000.
[7]D. T. Britto and H. J. Kronzucker, "Cellular mechanisms of potassium transport in plants," Physiologia Plantarum, vol. 133, no. 4, pp. 637-650, 2008.
[8] K.-H. Chan, W. Ke, and S.-K. Im, "CARU: A content-adaptive recurrent unit for the transition of hidden state in NLP," in Neural Information Processing: 27th International Conference, ICONIP 2020, Bangkok, Thailand, November 23–27, 2020, Proceedings, Part I 27, 2020: Springer, pp. 693-703.
[9]W. Chan, N. R. Ke, and I. Lane, "Transferring knowledge from a RNN to a DNN," arXiv preprint arXiv:1504.01483, 2015.
[10]C. Charalambous, "Conjugate gradient algorithm for efficient training of artificial neural networks," IEE Proceedings G (Circuits, Devices and Systems), vol. 139, no. 3, pp. 301-310, 1992.
[11]M. D. Clark, G. F. Contreras, R. Shen, and E. Perozo, "Electromechanical coupling in the hyperpolarization-activated K+ channel KAT1," Nature, vol. 583, no. 7814, pp. 145-149, 2020/07/01 2020, doi: 10.1038/s41586-020-2335-4.
[12]N. Creux and S. Harmer, "Circadian rhythms in plants," Cold Spring Harbor Perspectives in Biology, vol. 11, no. 9, p. a034611, 2019.
[13]X. Dai and R. Gretsch, "Quasi-synchronous sampling algorithm and its applications," IEEE Transactions on Instrumentation and Measurement, vol. 43, no. 2, pp. 204-209, 1994.
[14]J. Dainty, "Ion transport and electrical potentials in plant cells," Annual Review of Plant Physiology, vol. 13, no. 1, pp. 379-402, 1962.
[15]W. H. Drinkenburg, G. S. Ruigt, and A. Ahnaou, "Pharmaco-EEG studies in animals: an overview of contemporary translational applications," Neuropsychobiology, vol. 72, no. 3-4, pp. 151-164, 2016.
[16]M. A. Fattah and F. Ren, "GA, MR, FFNN, PNN and GMM based models for automatic text summarization," Computer Speech & Language, vol. 23, no. 1, pp. 126-144, 2009.
[17]J. A. Ferreira, "The multilevel modular DC converter," IEEE Transactions on Power Electronics, vol. 28, no. 10, pp. 4460-4465, 2013.
[18]M. Fried and R. E. Shapiro, "Soil-plant relationships in ion uptake," Annual review of plant physiology, vol. 12, no. 1, pp. 91-112, 1961.
[19]B. He, Modeling & imaging of bioelectrical activity: principles and applications. Springer Science & Business Media, 2010.
[20]P. Heckbert, "Fourier transforms and the fast Fourier transform (FFT) algorithm," Computer Graphics, vol. 2, no. 1995, pp. 15-463, 1995.
[21]J.-C. Henquin and H. Meissner, "Significance of ionic fluxes and changes in membrane potential for stimulus-secretion coupling in pancreatic B-cells," Experientia, vol. 40, pp. 1043-1052, 1984.
[22]A. L. Hodgkin and A. F. Huxley, "Resting and action potentials in single nerve fibres," The Journal of physiology, vol. 104, no. 2, p. 176, 1945.
[23]T. Hölttä, T. Vesala, S. Sevanto, M. Perämäki, and E. Nikinmaa, "Modeling xylem and phloem water flows in trees according to cohesion theory and Münch hypothesis," Trees, vol. 20, pp. 67-78, 2006.
[24]Y. Hu, A. Huber, J. Anumula, and S.-C. Liu, "Overcoming the vanishing gradient problem in plain recurrent networks," arXiv preprint arXiv:1801.06105, 2018.
[25]N. G. Irani and E. Russinova, "Receptor endocytosis and signaling in plants," Current opinion in plant biology, vol. 12, no. 6, pp. 653-659, 2009.
[26] C. Lea, R. Vidal, A. Reiter, and G. D. Hager, "Temporal convolutional networks: A unified approach to action segmentation," in Computer Vision–ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8-10 and 15-16, 2016, Proceedings, Part III 14, 2016: Springer, pp. 47-54.
[27]P. Newbold, "ARIMA model building and the time series analysis approach to forecasting," Journal of forecasting, vol. 2, no. 1, pp. 23-35, 1983.
[28] V. Nguyen and R. Nitisoravut, "Bioelectricity Generation in Plant Microbial Fuel Cell Using Forage Grass under Variations of Circadian Rhythm, Ambient Temperature, and Soil Water Contents," in 2019 IEEE Asia Power and Energy Engineering Conference (APEEC), 29-31 March 2019 2019, pp. 240-244, doi: 10.1109/APEEC.2019.8720344.
[29]A. Pal, A. K. Gautam, and Y. N. Singh, "Evaluation of bioelectric signals for human recognition," Procedia Computer Science, vol. 48, pp. 746-752, 2015.
[30]T. E. Rawlins, "Phytopathological and botanical research methods," (No Title), 1933.
[31]A. Saveyn, K. Steppe, and R. Lemeur, "Daytime depression in tree stem CO 2 efflux rates: is it caused by low stem turgor pressure?," Annals of Botany, vol. 99, no. 3, pp. 477-485, 2007.
[32]M. H. Sazli, "A brief review of feed-forward neural networks," Communications Faculty of Sciences University of Ankara Series A2-A3 Physical Sciences and Engineering, vol. 50, no. 01, 2006.
[33]A. Sherstinsky, "Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network," Physica D: Nonlinear Phenomena, vol. 404, p. 132306, 2020.
[34]T. Sibaoka, "Rapid plant movements triggered by action potentials," The botanical magazine= Shokubutsu-gaku-zasshi, vol. 104, pp. 73-95, 1991.
[35]D. Srivastava et al., "Role of circadian rhythm in plant system: An update from development to stress response," Environmental and Experimental Botany, vol. 162, pp. 256-271, 2019.
[36]B. Stanković and E. Davies, "Both action potentials and variation potentials induce proteinase inhibitor gene expression in tomato," Febs Letters, vol. 390, no. 3, pp. 275-279, 1996.
[37] M. Ventresca and H. R. Tizhoosh, "Opposite transfer functions and backpropagation through time," in 2007 IEEE Symposium on Foundations of Computational Intelligence, 2007: IEEE, pp. 570-577.
[38] R. Zhang, Z. Yuan, and X. Shao, "A new combined CNN-RNN model for sector stock price analysis," in 2018 IEEE 42nd Annual Computer Software and Applications Conference (COMPSAC), 2018, vol. 2: IEEE, pp. 546-551.
[39]S. Zimmermann, T. Ehrhardt, G. Plesch, and B. Müller-Röber, "Ion channels in plant signaling," Cellular and Molecular Life Sciences CMLS, vol. 55, no. 2, pp. 183-203, 1999/02/01 1999, doi: 10.1007/s000180050284.