Artificial Neural Network Application Technology Research In The Measurement Of Heat

Steam is a kind of important secondary energy in industry. It is important to measure steam flow accurately for economizing energy and advancing the benefit of the enterprise. However, being a kind of special carrier, steam is different from other gases, so it is difficult to carry out its measurement.In this dissertation, a Artificial Neural Network(ANN) model is introduced to replace the traditional algorithm in the data process of steam flow measurement. It contains: how to select the ANN model, how to build and optimize the ANN model , the way to improve BP algorithm ,the analysis of filter characteristic of BP model , and how to realize the ANN model in the microprocessor.The microprocessor we prefer to is PIC16C74A, designed and produced by Microchip Company in U.S.A.. It adopts RISC technology, and have the advantage of powerful function, easy program ,and simple periphery circuit.Conclusions of this dissertation show that the ANN model have the advantage of simple algorithm, fast operation, saving memory room , reducing noise . Adoption of ANN model in secondary instrument helps to increase its integrate degree.