| As one of the main facilities in the agricultural production process,the solar greenhouse is composed of thermal insulation and heat storage walls,high light transmission film and certain enclosure structure,which can ensure the temperature of the greenhouse at night through the heat storage and release of the back wall,so as to achieve the purpose of overwintering and cultivating the growth of crops,which is an important reflection of the level of development of modern agriculture in a country.With the in-depth study of solar greenhouse,agricultural researchers have realized that the temperature inside the greenhouse is one of the main reasons affecting the yield and quality of greenhouse crops,and either too high or too low temperature is not conducive to the normal growth of crops.Therefore,it is especially important to explore the spatial distribution pattern of temperature inside the heliostat and establish a temperature gradient distribution model to reasonably regulate and improve the growth environment of greenhouse crops in winter,which is of great significance to the steady growth of agriculture and farmers’ income.Based on the existing research on the temperature of sunlight greenhouse,this paper takes the soil-back wall solar greenhouse of Northwest Agriculture and Forestry University Horticultural Farm in Yangling,Shaanxi Province as the research object,and analyzes the daily variation pattern and spatial distribution pattern of the temperature of the test greenhouse by collecting indoor temperature and outdoor meteorological data.On the basis of this study,a co-integration theory and error correction model are introduced to establish a co-integration theory-based indoor average temperature prediction model and a spatial temperature distribution model,and the validity of the model is verified by comparing the model calculated values with the actual values,and draw the following main conclusions:(1)Linear regression analysis method is a commonly used method for predicting temperature in solar greenhouses.This study proposes an improved method for pseudoregression problems that may occur in the prediction process,that is,the model can be improved by introducing lag variables to eliminate residual errors.autocorrelation to establish an error correction model.(2)The trend of indoor temperature change with outdoor temperature is basically the same in the earthen back wall solar greenhouse,which generally shows a trend of ’one up and two down’.Solar radiation is the most important factor that causes the change of indoor temperature,no matter it is sunny or cloudy,the indoor temperature is proportional to the intensity of outdoor solar radiation.When the insulation quilt is uncovered,the indoor temperature rises rapidly,reaching the maximum temperature around 12:00 on cloudy days and 13:00 on sunny days,and then the temperature will drop rapidly and become stable after covering the insulation quilt.(3)Regardless of sunny and cloudy days,the spatial distribution pattern of temperature in the heliostat was basically the same.The temperature gradually increased from west to east,with the maximum temperature difference of 3.0℃ on sunny days and 2.0℃ on cloudy days.At the same height,the nighttime north-south temperature showed a trend of gradually decreasing temperature from north to south,and this change was small,with the temperature difference not exceeding 1.0℃.The daytime north-south temperature can be roughly summarized as south high and north low,and this trend is most obvious under clear sky condition,and the temperature in the south and central part of the greenhouse is significantly higher than that in the north.Vertical temperature distribution is more complex,different weather and time,the distribution trend are different,the night is basically uniform distribution,vertical temperature difference is very small;sunny daytime,due to the influence of solar radiation,its indoor temperature from the ground up gradually,but when the height of more than the back wall of the vent,the temperature began to fall again.During the daytime on cloudy days,the overall vertical direction showed a trend of gradually increasing from low to high temperature,but the change was very small,and the maximum temperature difference between the two end points was only 0.5℃.(4)In this study,an indoor temperature prediction model based on the covariance theory was developed through the relationship between outdoor meteorological data and indoor temperature,which can achieve the forecast of the average temperature inside the heliostat using outdoor meteorological environmental factors such as outdoor temperature,humidity,previous day’s minimum and maximum temperature and solar radiation.The results showed that the prediction accuracy of the error correction model was higher and significantly better than that of the commonly used traditional linear regression model.The root mean square error and the mean absolute error of the error correction model under cloudy conditions were 0.42℃and 0.36℃,respectively,which were 16.0 and 14.0 percentage points lower than those of the traditional linear regression model of 0.58℃ and 0.50℃,respectively.The root mean square error and the mean absolute error of the error correction model under sunny conditions were 1.59℃ and 1.36℃,respectively,which were 35.0 and 44.0 percentage points lower compared with 1.95℃ and 1.80℃,respectively,of the traditional linear regression model.It indicates that the co-integration theory-based heliostat temperature prediction method proposed in this paper has high prediction accuracy respectively,which are 35.0% and44.0% lower than the 1.95℃ and 1.80℃ of the traditional linear regression model.The prediction method of greenhouse temperature based on the theory of co-integration has high accuracy.(5)In this paper,a model of spatial temperature distribution in a heliostat was established by analyzing the change pattern of indoor spatial temperature distribution.By comparing the measured and calculated values at 9:00 and 13:30 under typical sunny weather,it is concluded that the model has the highest prediction accuracy for 9:00,with the maximum absolute error of 0.69℃ and root mean square error of 0.37℃;the prediction effect for 13:30is slightly worse,with the maximum absolute error of 1.30℃ and root mean square error of 0.73℃.In general,the temperature value of a point on the space of heliostat obtained by the model is more accurate and can be used for the preliminary analysis and calculation of greenhouse space temperature.To sum up,the indoor average temperature prediction model and spatial temperature distribution model proposed in this paper enrich the greenhouse temperature measurement system and have high prediction accuracy,which can provide a certain decision-making basis and reference significance for the temperature regulation in the solar greenhouse. |
PDF Full Text Request