基于MEA-LM-BP神经网络的棚室温度预测模型研究

Research on Greenhouse Temperature Prediction Model Based on MEA-LM-BP Neural Network

对棚内温度进行精确预测是实现棚室精准调控的前提,是实现棚内作物高品质栽培的保障。因棚室具有大惯性、强耦合、非线性等特点,通过机理分析法难以建立其准确的数学模型,人工神经网络方法在棚室温度预测方面应用广泛,但其存在的收敛速度慢、容易陷入局部最小等缺点使测量精度受到影响。为进一步提高基于神经网络算法的棚室温度预测模型精度,运用思维进化算法(Mind Evolutionary Algorithm,MEA)优化LM-BP神经网络模型的输入权重和阈值,并与遗传算法优化LM-BP网络模型和LM-BP模型进行对比。试验结果表明:MEA-LM-BP与LM-BP和GA-LM-BP方法相比,RMSE分别降低了0.32和0.16,平均相对误差降低了1.08%和0.58%。该方法提高了基于神经网络算法的棚室温度预测模型精确度,并提供了新的优化路径。

Accurate prediction of temperature in the greenhouse is the premise of accurate control in the greenhouse and the guarantee of high quality cultivation of crops in the greenhouse. Due to the large inertia, strong coupling and non-linear characteristics of the greenhouse, it is difficult to establish an accurate mathematical model through the mechanism analysis method. The artificial neural network method is widely used in the temperature prediction of the greenhouse, but its disadvantages such as slow convergence rate and easy to fall into the local minimum affect the measurement accuracy. In order to further improve the accuracy of prediction model of greenhouse temperature based on neural network Algorithm, this paper USES Mind Evolutionary Algorithm(MEA) to optimize the input weight and threshold of lm-bp neural network model, and compares it with genetic Algorithm to optimize lm-bp network model and lm-bp model. The test results showed that compared with lm-bp and ga-lm-bp methods, RMSE decreased by 0.32 and 0.16, respectively, and the average relative error decreased by 1.08% and 0.58%. This method improves the accuracy of temperature prediction model based on neural network algorithm and provides a new optimization path for it.