Studies On The Automatic Regulating System For Rice Water-Saving Irrigation In South Hilly Region

With an increasingly tense of water resources, every country in the world is actively exploring effective ways for water saving. Agriculture is the largest water consumer and the 2/3 fresh water in the world is consumed by agricultural production. Developing water-saving irrigation is very important for water saving. Rice is the basic fine grain crops, its cultivation area and production is just second to that of wheat in the world. In China, the ratio of rice area is 28% of the total grain area, whereas, its water consumption gets up to about 54% of the total water consumption, more than 65% of the total agriculture water consumption. So improving rice's water-saving irrigation technology & implementing scientific irrigation management is of great importance for the food safety and agriculture sustainable development of our country.Some advanced technologies such as "3S" technology, information technology & automatic control technology etc., are current hot topics of water-saving irrigation. The southern hilly area is an important agricultural production base in our country, most of them are small area & irregular terrace and with abundant rainfall but there is seasonal drought. Therefore, this thesis researched and developed a high- efficient automatic control & management system for rice water-saving irrigation in the light of the characteristics of southern hills region. The field test, physical production, statistics analysis and artificial intelligence are adopted in the research and achievements are as follows.1) Analyzed the soil moisture content spatial variation of hill region with the help of statistics and geo-statistics, then designed three sample solutions including queue array, ray array and asterisk array according to the spatial variation characteristics of rice field moisture content. Further verified the three sample solutions with the way of t-test. The analysis results indicate that the paddy soil moisture is normal distribution and the spatial variation coefficient is 11.88%, the variation is small, the ratio of gold value and the base value is 79.23%, the space correlation of moisture content is weak, the paddy field's moisture spatial distribution is heterogeneous. At the same time, the t-test results show that the errors of the three sample solutions are less than 3%.2) Developed a wireless-based field monitor system according to the characteristics of hilly areas and TDR soil moisture sensor are used. At the same time, a calibration of TDR soil moisture sensor is conducted for the rice soil and related performance test is carried out with the self-made wireless monitor system. The calibration test results shows that the relationship between y, the moisture content of rice soil, and x, the output voltage of sensor is y=49.553x+4.705. The field measure instrument is of small, low power consumption, low cost and easy operation. The performance test indicates that the field monitor system's energy saving is well, measurement error is less than 5% and transmission accuracy rate within 1000m is 100%.3) Designed a fuzzy automatic control system of rice water-saving irrigation for the hilly region. It is in the field of the irrigation index of water depth or moisture content and used the fuzzy control theory. Established a "2-inputs and single-output" fuzzy controller, taking the proportion value of soil moisture content divided by set value and its rate as inputs, the water pump irrigation time as output. The simulation results show that the automatic control system is stable and it can save water about 10% compared with artificial irrigation.4) Developed a terminal controller for the automatic irrigation system, implemented the communication with PC, the pump operation parameters setting, and the irrigation information preservation and displaying during the water pump process. Therefore, it provided conditions for the irrigation automation management in rural area.5) Formulated a nonlinear relationship between It, rice field irrigation amount,Pt, the rainfall,and St, irrigation area by the field water balance formula and established two forecast models for the rice irrigation using BP neural network.â‘ Taking the rainfall, irrigation area St as the inputs and the rice field irrigation amount It as the output, established a "2-4-1" BP neural network forecast model.â‘¡Taking the three rainfall Pt-3,Pt-2,Pt-1 prior to t-Period,the irrigation area St as inputs and the rice field irrigation amount It as the output, established a "4-6-1" BP neural network forecast model. Furthermore, taking the three rainfall Pt-3,Pt-2,Pt-1 prior to t-Period as inputs and the rainfall Pt as output, established a "3-5-1" BP neural network rainfall forecast model. Applied 18 years' historic data of the test field to test the effectiveness of the forecast models, the results indicated that prediction accuracy of 1st prediction models is more than 95% and that of 2nd prediction model is about 90% and the prediction accuracy of the rainfall forecast is more than 90%. It provided the basis for implementing dynamic water distribution and provided the effective way for solving the hilly seasonal drought.6) Designed the software for rice water-saving irrigation automatic control system, and then developed a rice automatic water-saving irrigation system which integrated the real-time monitor, prediction and automatic irrigation.This study developed an automatic regulation management system of hilly-region rice water-saving irrigation. It integrated the real-time monitor, prediction and automatic irrigation, provided an application model of precision irrigation and some information management ideas for rice water-saving irrigation in China rural area.