A fuzzy inference system was developed for predicting the heat index from temperature and relative humidity data. The effectiveness of fuzzy logic in using imprecise mapping of input to output to encode interconnectedness of system variables was exploited to uncover a linguistic model of how the temperature and humidity conditions impact the heat index in a growth room. The developed model achieved an R2 of 0.974 and a RMSE of 0.084 when evaluated on a test set, and the results were statistically significant (F1,5915 = 222900.858, p < 0.001). By providing the advantage of linguistic summarization of data trends as well as high prediction accuracy, the fuzzy logic model proved to be an effective machine learning method for heat control problems.
翻译:为了从温度和相对湿度数据中预测热指数,开发了一个模糊的推论系统,利用模糊逻辑对输出输入进行不精确的绘图以编码系统变量的相互关联性,发现温度和湿度状况如何影响生长室热指数的语言模型,开发的模型在用一套测试来评价热指数时达到0.974R2和0.084RME,结果具有统计意义(F1,5915=222900.858, p < 0.001)。通过提供数据趋势语言汇总的优势以及高预测准确性,模糊逻辑模型证明是处理热控制问题的有效机器学习方法。