Research On The Air Preheater's Hot Spot Detection System Based On The Infrared Array Sensor

This paper mainly studies the hot spot detection system of air pre-heater, which based on the infrared array sensor, as well as the data fusion algorithm to match with the system.Air preheater is the device that uses the waste gas of the boiler to heat the air which can be used again in the boiler. During shutdown or low-load operation, due to the lower furnace temperatures caused by incomplete combustion of some flammable material and deposited in the heat transfer surface. These sediments are prone to spontaneous combustion, which is known as the secondary combustion air preheater. The need for air preheater's hot spot detection to prevent them is necessary. At present, most of the air preheater hot spot detection systems are based on thermocouples or infrared sensors as a temperature measurement device. Compared temperature value with the pre-set alarm threshold to the value as a way to determine whether the need for alarm. However, air preheater's fire may be occur within the temperature range, it is difficult to determine an appropriate temperature threshold value to detect all occurrences of the fire, and the threshold set too much depends on experience and on-site working conditions, prone to leakage alarm and false alarm. On the other hand, due to a single infrared sensor data is isolated and can not be timely refleced the temperature of the air preheater state, so the system put forth an array of data fusion technology based on infrared fire detection and alarm methods. Many infrared sensors detect the air preheater. According to the temperature of the air preheater distribution of these data, the system uses D-S evidence theory to integrate effectively.The system consists of the upper computer and Lower computer. Each infrared probe consists of a few infrared sensors which mounted on two separate MCU. Air preheater has a number of infrared probes. Through dual bus structure, the host computer for real-time communication uses IPC data processing and final decision-making. When the system began to run, the system can detect the value of multi-sensor, and find the internal temperature distribution, based on self-regulation, which can continue to detect the hot spot temperature's location and size of hot spots. In addition, this system also has a certain self-diagnosis and self-adaptive ability. When a sensor fails. it can automatically adjust fusion algorithm according to specific fault conditions. then issue a corresponding alarm signal.