According to the data you provided, there is a simplest mathematical model called seasonal index method. The basic operation method is as follows: find the monthly average stock of 20 1 1 year, and then divide the monthly average stock of 20 1 1 year by the average value to get the seasonal index of the month. Then divide the data of the known month of 20 12 years (for example, 1 month) by the current month (seasonal index of 1 year) to get the average inventory value of 20 12 years. Then multiply the seasonal index of each subsequent month by the average inventory to get the monthly forecast value. This is a method, but it requires that the stock changes have obvious seasonal cycle characteristics. The second method is exponential smoothing. This method requires less data. But it also has its limitations. First, the smoothing index should be set by experience to see whether the stock in your warehouse fluctuates violently. Second, you can only make a very short-term forecast (that is, a calculation cycle, according to your situation, it can only be May. Only when the measured value in May is available can we predict the measured value in June.