應用Thermopile 感應器所設計的溫度測量裝置，生產時都必須作好熱敏電阻及Thermopile感應零敏度的校正。免熱敏電阻校正的Thermopile 感應器演算法，用於改善舊有以Thermopile感應器為測量元件所製作的溫度量測裝置，在生產時所面對的複雜的校正問題。主要目的為提高測量準確度及大幅降低校正所需的時間，由原來需要5分鐘的校正時間，改善為4秒鐘完成校正。係應用Thermopile的Cold Junction與Hot Junction的溫度差異關係，對應Thermopile輸出的電壓曲線，分析其相關特性，確定曲線符合數學模式，可以建立一組聯立方程式，並解出相對應的兩組重要參數(熱敏電阻誤差及Thermopile感應零敏度)。達到不需要單獨針對熱敏電阻作個別校正的目的。 大多數以Thermopile Sensor 所設計的測量裝置，都以8-Bit MCU為運算處理單元，無浮點運算能力。解聯立方程式極為困難。為了解決這個問題，在本文中探討以一種8-Bit MCU可執行的循環運算流程，達到解聯立方程式的運算目的。這項運算流程同樣適用於部份微處理器在代數運算方面的應用。 Conventional thermopile thermometers must perform two calibration procedures for the thermistor tolerance and the thermopile sensitivity during the production. This project focuses on the calibration-free algorism for thermistor tolerance of thermopile sensor. It intends to simplify the complex calibration process of thermopile thermometers. The aim is to increase the measurement accuracy and reduce the calibration time. By analyzing the characteristic of the thermopile output voltage curve that is relevant to the temperature difference between the cold junction and hot junction, a mathematical model that fits the curve can be found. Therefore, we can form a simultaneous equation and thereby solve it to obtain two essential parameters: thermistor tolerance and thermopile sensitivity. As a result, there is no need to calibrate thermistor individually. By applying this algorism, the calibration time can be reduced from 5 minutes to 4 seconds. Most of thermopile thermometers are implemented with 8-bit MCUs as the central processing unit. However, this type of MCU does not have the floating point unit to execute a complex mathematical calculation such as to solve simultaneous equations. To solve this issue, this project applies a particular calculation flowchart that can be used by an 8-bit MCU to solve simultaneous equations. This calculation flowchart is also suitable for other MCUs that need to solve complex mathematics.