Research on the requirement planning of APS production scheduling software considering time sequence

Other features

The implementation requirements plan requires that the special requirements of the partners working together in the supply chain be clearly articulated. Therefore, it is necessary to understand the time series you are considering. Because the characteristics of a sequence are different from the predicted dimensions (time, product, geographic range), different prediction processes must be used.

Occasional demand

If there is no demand for a considerable period of time, we call this sequence of events sporadic (intermittent). Those requirements patterns are particularly present in the replacement requirements section or only in a small portion of the forecast demand, such as the need for a particular size of denim clothing that may be sporadic in a store one day. The use of ordinary statistical forecasting methods for these products can make a great error. Other predictive methods also fail to improve the quality, since the appearance of the demand-free period is usually purely random, and therefore unpredictable. And occasional demand often appears in a large number of C-class products, for these products, using the Planner's low-cost, low http://www.aliyun.com/zixun/aggregation/18194.html "> Time consumption advantage to get the forecast is worthwhile."

Therefore, an effective program for automatic calculation of this prediction is developed for the prediction of accidental demand. These methods try to predict separately the emergence of two component "positive (positive) demand periods" and "demand". For example, determine the interval between two transactions and the transaction volume. The update of component data is accomplished by simple exponential smoothing method. If the contingency requirements process does not produce the specific impact that can generate the intermittent demand model, it is possible to have a large number of less observed errors. For example, the presence of frequent out-of-stock in a retail store generates a time series that implies occasional demands.

Comparison of Unsalable and delayed orders (back korders)

Demand forecasting is usually based on the historical data of the product. However, industrial customers (business-to-business) often accept deferred orders, and customers will not accept them if they do not get the product in time. As a result, the total amount of sales we observe equals the total demand in the case of an unpaid order, but the sales value in a unsalable situation is likely to underestimate the real demand. In order to predict, a requirement time series needs to be used, so the demand time sequence must be calculated from the observed sales value. This type of problem occurs frequently if you want to calculate demand forecasts for points of sale (retailer, store).

There are basically two different ways to solve the demand forecasting problem when there are unsalable. The first method tries to calculate the historical information of the actual demand, which is based on the historical information of the sales and the information of the Out-of-stock. Demand forecasting is calculated on the basis of the historical information of actual demand. If the number of Out-of-stock is very small, this method will produce better results. Another alternative to the problem of unsalable is the use of empirical statistical methods, which take the observed demand values as sample samples of requirements. To use this method you must understand the inventory management process, which is applied to the product you are considering.

Yongkaida APs, to solve the scheduling problem that cannot be satisfied by ERP Http://www.yukontek.com,APS Production planning management expert 400-076-7600,021-68886010