The Design Of Embedded Controller Used For Acid Gas Waste Heat Recovery Device

With the rapid development of industrialization, boilers have been widely used to various fieldsof production and life. Nowadays, the volume of the boiler is getting bigger and bigger and muchmore complex structure, the degree of automation is also increasing. Boilers make a huge contributionto the economy, at the same time, because of the low thermal efficiency, high waste heat dischargetemperature, inevitably bring about serious environmental pollution and energy consumption,therefore, recycle flue gas waste heat has important practical significance, this paper main focuses onthe embedded controller used in the flue gas heat recovery unit. The main contents are as follows:First, introduce our country energy consumption and environmental pollution this situation dueto industrial boilers low thermal efficiency, under the background of growing energy shortage, theimportance of recycling flue gas heat recovery is self-evident. Home and abroad have developed theheat recovery unit, but the mainstream products are based on PLC controller, which covers a largearea and complex system construction. The embedded controller can solve this problem.Secondly, make a detailed work and design ideals introduction about the waste heat recovery unitbased on embedded controller S3C2440, including hardware circuit and control algorithms:temperature acquisition and signal conditioning circuit based on temperature sensor PT100; alarmcircuit design based on pressure sensor PT124B-111; motor control circuit based on opt coupler;recovery boiler operating parameters short message sending module based on the GSM; multi-channelA/D conversion; PWM control, and pass all of the features through online debugging.Finally, introduce the characteristics of boiler failure and common boiler fault diagnosis methods,the basic principle of neural network and wavelet transform. Based on wavelet transform and BPneural network, this paper uses wavelet neural network diagnosis and predict boiler failure, theexperimental results show that this method is effective to avoid the slow convergence of BP neuralnetwork and its easy to fall into local minimum point, wavelet neural network has much more faultdiagnosis capability than BP neural network.