CAJ | 학술논문

针对现有盆栽系统难以根据植物类型定制个性化浇灌方案、水资源利用率不高等问题，综合利用物联网、时间序列分析技术与规则库对家庭盆栽养植的智能化自动浇灌系统进行研究。利用传感器采集土壤信息和环境信息，并构建带有外因输入的自回归移动平均模型对未来土壤湿度进行预测，将预测结果结合专家库中相关规则对是否浇灌进行决策，利用历次用水量数据和环境变量预测当前用水量并完成自动浇灌。通过实验分析验证，该系统能够使盆栽植物的土壤湿度保持在适宜生长范围内，避免剧烈变化，同时通过预测方法实现对盆栽的精准浇灌，达到节水的效果。
침대현유분재계통난이근거식물류형정제개성화요관방안、수자원이용솔불고등문제，종합이용물련망、시간서렬분석기술여규칙고대가정분재양식적지능화자동요관계통진행연구。이용전감기채집토양신식화배경신식，병구건대유외인수입적자회귀이동평균모형대미래토양습도진행예측，장예측결과결합전가고중상관규칙대시부요관진행결책，이용력차용수량수거화배경변량예측당전용수량병완성자동요관。통과실험분석험증，해계통능구사분재식물적토양습도보지재괄의생장범위내，피면극렬변화，동시통과예측방법실현대분재적정준요관，체도절수적효과。
At present, irrigation systems for potted plants are unable to adjust personalized irrigation schemes according to the types of plants, and with a low utilization of water resource. For the above problems, an intelligent irrigation system was developed for the family potted plants using the technology of Internet of things, time series analysis and rule base. First, the system used sensors to collect information of soil and environment, and then built autoregressive moving average with exogeneous inputs model ( ARMAX model) to forecast the soil moisture; second, the system combined the predicted results with related expert rules to make decisions, and then decided whether the system should irrigate or not; third, the system used the previous water consumption data and environment variables to predict the current irrigation water consumption, and then it completed the automatic irrigation. Through the experimental analysis, the system can keep the soil moisture of potted plants changing smoothly in a suitable range, at the same time, through the predicting method, the system can irrigate the potted plants precisely and achieve the effect of water saving.