Font Size: a A A

Research On Plant Information Perception And Self-organized Agricultural Internet Of Things System

Posted on:2013-02-02Degree:DoctorType:Dissertation
Country:ChinaCandidate:P C NieFull Text:PDF
GTID:1118330371956327Subject:Agricultural Electrification and Automation
Abstract/Summary:PDF Full Text Request
As one of the forefront areas in the modern agriculture, the agricultural internet of things (IOT) is a key and kernel technology in the development of agricultural informatization and sustainable realization of the world. The Agricultural IOT information technology mainly includes three levels which are agricultural information perception, transmission and information application. While the traditional farmland information acquisition faces several major technical bottleneck:one is backward technology of sensors, as a result, much information of crops lacks corresponding sensors. In addition, the traditional farmland information monitoring is just the single point and static timing measurement which is unable to realize real-time dynamic detection, let alone realize the request of unattended agricultural automation. Therefore, there are some crucial problems that await to be solved urgently in modern agriculture. One is to study the technology of the key sensors and instruments, the wireless transmission agreement and depth routing mechanism based on the characteristics of agricultural IOT netting; the other is to develop the hardware and software platform of the agricultural IOT. Three core levels of agricultural IOT were the objects of this study, nondestructive determining technology of plant nutrients was discussed based on visible/near infrared spectra, and the weak signal detection and development of the corresponding sensing instruments were presented based on the FFT and wavelet transformation. In the research, it was proposed that the active induced self-organized network protocol on large-scale agricultural IOT and the depth routing technology of agricultural IOT. Also,it was investigated that how the intelligent routing maintained under fault conditions, and instruments was developed on the information collection and control systems which had already been successfully used in the agricultural production. The main creative contributions of this paper include:(1) A determining method of plant nutrients was proposed based on visible/near infrared spectra. Through analysis methods of spectral data preprocessing-characteristic wavelengths selection-linear and nonlinear calibration modeling,13 effective wavelengths were extracted for the detection of plants nutrients, while 3 effective wavelengths were extracted for plant canopy nutrient information detection, and some sensors were developed for real-time and dynamic detection of this information. The tests of these sensors showed that R2=0.8237 for Nitrogen content detection, R2=0.9361 for chlorophyll detection, R2=0.9672 for NDVI detection, while R2=0.7698 for LAI detection. Some analytical mothods were established for the simultaneous determination of single point chlorophyll content, nitrogen content and moisture content in leaves. Some instruments were developed, and the results of tests were as follows:R2=0.9148 for chlorophyll content detection, R2=0.9207 for nitrogen content detection, and R2=0.8656 for moisture content detection.(2) Photoelectric sensors with high-sensitivity photoelectric micro-signals outputting were applied as the spectral information detector of plant nutrients, signal circuit was designed for processing, and processing methods were investigated for output signal. Filter and weak signal extraction were done by FFT algorithm and wavelet transform analysis, respectively. Results showed that signals processed by Wavelet transform were more close to the original useful signal, it was approximately the same of FFT algorithm and wavelet transform analysis in high frequency band, and de-noising effect by wavelet transform analysis was a little better in low frequency band. As a result, the amplitudes of noise dropped below 0.5uv, while SNR (signal-to-noise) was improved to 8db.(3) Self-organized network protocol of active induced agricultural IOT was discussed adapted to large-scale agricultural information collection, the realization method was proposed based on F-MSG, B-MSG message event-driven system networking. In this foundation, some optimal management methods were presented as the shortest path routing optimization of self-organized network protocol (SLR), optimization of routing depth defense(S-SLR), optimization mechanism of competition mode(X-SLR). And experiments showed that QoS was 1.13 for SLR, QoS was 0.14 for X-SLR, QoS was 0.23 for S-SLR. Network performances were improved. In this foundation, information scheduling of routing depth and networking management mechanism were presented, routing depth of Agricultural IOT leveled up to 12, other parameters were the same except that network delay was slightly increased as QoS=0.83. The fine network performances meet the practical demands of agricultural industry.(4) Intelligent routing maintaining method as responses to network nodes faults was presented by self-organized network protocol technology. Two methods were proposed based on local network reorganization and leapfrog routing, and experiments showed that average network delay was below 5 seconds by routing maintenance of local network reorganization, packet loss rate of network was well controlled below 1.5%,Qos was 0.15, and packet loss rate of network by leapfrog routing maintance was controlled below 3%, average network delay was 8 seconds,QoS was 0.26. It had been proved that these two routing maintenance methods completely satisfied with the desired network performances of the agricultural IOT.(5) In the context of real-time agricultural IOT information acquisiton, intelligent information management system of agriculture was proposed of agricultural IOT system combined with agricultural automatic control devices. The methods for constant pressure control of pump and automatic management of fertilizer and water were discussed, and some corresponding systems were developed and successfully applied in plantation as demonstration.The studies above have laid a theoretical foundation for real-time dynamic acquisition of multi-dimensional information and intelligent low-power consumption of remote transmission and automatic control in large-scale field, and it has an extensive application prospect.
Keywords/Search Tags:agricultural IOT, spectral detection technology, weak signal detection, agricultural sensing instruments, active induced networking protocol, routing depth of IOT, automatic control of agriculture
PDF Full Text Request
Related items