As the foundation of the country and people's livelihood, the price stability of agriculture and food is of great significance to the social and economic development. In recent years, the prices of agricultural products are affected by weather, transportation and policies, and the prices are characterized by non-linear changes, which seriously affects the social stability and people's lives. In order to accurately predict the price of eggplant, this paper proposes a combination model based on CNN_LSTM. Firstly, we use Spearman correlation coefficient (Spearman) to derive the correlation coefficient between meteorological factors and prices to determine the main meteorological factors, secondly, we use CNN convolutional layer to extract the dynamic characteristics of the meteorological factors and price data, and then again, we use LSTM layer to extract the characteristics of the input time-series data, and then we complete the price prediction and output through the output layer. In this paper, the historical meteorological and eggplant price data of Santai County, Sichuan Province in the past 10 years are used, and the CNN, LSTM and CNN-LSTM models are compared and analyzed, and the results show that the RMES of CNN-LSTM model is 0.01, the MAE is 0.06, and the R2 is 0.99 is better than that of CNN and LSTM, and the predicted value and the actual value have a high degree of fit, and it can be used for the following purposes Agricultural price prediction, agro-meteorological economy, rural revitalization and policy regulation provide mindfulness and new methods.