Research On Neural Network Prediction Algorithm For Mineral Processing Cost

At present,Chinese had many concentrating plants,which are automated and costly to produce.In the past cost management,the cost analysis was not in line with the development concept of modern enterprises after the products were formed.The cost forecast beforehand is forward-looking and can help companies effectively reduce wear and tear.In this paper,the neural network algorithm model suitable for mineral processing cost prediction is designed for the calculation of mineral processing cost.The compilation of cost prediction software is completed by C#,which provides scientific theory for cost scientific management and subsequent cost control of small and mediumsized mineral processing enterprises.The mineral processing cost prediction algorithm uses BP neural network as the main prediction algorithm.The structure of BP neural network structure is discussed from several aspects by using grey relational analysis method and golden section optimization method.The ore dressing cost prediction software is compiled using the C# language through the Visual Studio environment.The software sets up the human-computer interaction interface and writes the BP neural network algorithm.The NPOI technology is used to read the data input in the Excel spreadsheet to the neural network to obtain a cost prediction value,which is presented to the user via the human-computer interaction interface.In addition,the software also integrates functions such as algorithm parameter viewing and historical data query to facilitate professional technicians and staff to view relevant data.After platform testing,the software is practical and stable,and has achieved the expected results.Figure 28;Table 6;Reference 62.