| According to China’s 2016 Water Resources Bulletin,China’s total water consumption is604.02 billion m~3,of which agricultural water consumption is 376.80 billion m~3,accounting for62.4%of total water consumption.At present,the effective water use coefficient of agriculture in China is only 0.52,while the level of other advanced agricultural countries is 0.7~0.8,which is a big gap.Therefore,it is very necessary to improve the utilization efficiency of water resources in agricultural production to alleviate the shortage of water resources in China.To solve this problem,the concept of“precise irrigation”was proposed by agricultural workers.Precision agricultural irrigation technology is based on field cultivation.According to the requirements of crop growth process,the most precise irrigation facilities are used to strictly and effectively fertilize crops to ensure the growth of crops.Therefore,the accuracy of irrigation is very important,and the accuracy of irrigation depends on the accuracy of soil moisture information collection.At present,the most common method used in production is to measure soil moisture content by sensor.However,there is a lack of performance evaluation of various sensors in the market,so it is difficult to distinguish the advantages and disadvantages of different sensors.Moreover,sensor manufacturers usually develop in the laboratory environment when designing and producing instruments,and the built-in calibration model cannot cope with the actual environmental changes,so that the performance claimed by manufacturers cannot be achieved in actual use,resulting in low production efficiency in practice,and the goal of improving agricultural water efficiency cannot be achieved.Therefore,the performance of sensors is compared under the same standard,and the influence of environmental changes on sensors is studied.At the same time,the sensors are compensated and corrected to improve the accuracy of soil moisture content information obtained in agricultural production.Improving the accuracy of obtaining soil moisture content information is undoubtedly a very effective solution for improving agricultural water efficiency and reducing water resource waste.Therefore,this study is mainly divided into the following three parts:(1)Select the soil moisture sensor commonly used in the market(the number is a,b,c,d in the test)for field precision test and indoor precision test to evaluate the accuracy of different sensors under different conditions.The test results in this part indicate that when the sensor faces different environmental changes,the soil measurement results for the same soil moisture content will have different measurement results as the external environment changes,and the measurement error value will also change,and Under the condition,the variation of the error of the sensor is quite obvious.Although the sensor manufacturer has made preliminary compensation correction for the external environment change during the sensor manufacturing process,it still has limitations compared with the real natural environment change.From the results of this test.It is important to recalibrate the sensor for specific environmental conditions if a more accurate measurement is desired.(2)Select the sensor used in the accuracy evaluation and conduct temperature influence test to study the influence of temperature change on different sensors and the correlation law of temperature influence sensor monitoring results.The experimental results in this part indicate that the sensitivity of different types of sensors to soil temperature is affected by other environmental factors and the sensor’s native differences,but the sensitivity of the sensor to temperature is independent of the sensor type.From the test results,the minimum effect of temperature on the sensor is the test result of the soil sample of the sensor b with a water content of 27.01%,and the total temperature change of the test is 3.30%.The maximum value of sensor c for soil moisture is 27.01%soil sample test,and the total test temperature change is14.70%.The effect of temperature on the sensor cannot be ignored.(3)Using the mathematical software method to compensate the sensor for correlation,compare the compensation results of different correction methods,and analyze the advantages and disadvantages of different correction methods.The experimental results of this part indicate that the test data of sensor b is selected,and the compensated water content prediction model is fitted by the binary regression method and the improved neural network method.The fitting results show that both the binary regression method and the neural network method can significantly reduce the deviation of the sensor monitoring data and reduce the influence of temperature on the sensor monitoring results.The water content prediction model and the binary regression method fitted by the neural network method Compared with the prediction model,the prediction model is more accurate and the deviation value is lower.The above conclusions indicate that when the sensor faces different environmental changes,the soil measurement results for the same soil moisture content will have different measurement results as the external environment changes,and the measurement error value will also change accordingly,which is difficult to reach the manufacturer’s claim.Work accuracy.Moreover,the influence of temperature on the working efficiency of the sensor cannot be ignored.Both the binary regression method and the neural network method can significantly reduce the deviation value of the sensor monitoring data and reduce the influence of temperature on the sensor monitoring results.The water content prediction model fitted by the neural network method and the prediction model fitted by the binary regression method are More accurate,and the bias value is lower.Using the mathematical software method to compensate the sensor can reduce the error of the sensor work and improve work efficiency. |
PDF Full Text Request