Leakage of organic gas in an industrial environment can easily cause fire and explosion accidents. Rapid detection of gases type and concentration is the premise for emergency treatment of these situations. Typically, the single semiconductor sensor cannot detect multicomponent gas. Here, a new multicomponent gas detection method was proposed based on the temperature-response relationship of the sensor. This method has low power consumption, low cost, and long-term monitoring, which can guide the future exploration of deployable sensors. First, the parameters were optimized according to cosine distance to reflect the maximum difference in the temperature-response relationship. Then coordinate system transformation was carried out to further magnify the difference between gases. Finally, the multicomponent gas was recognized by using rational Taylor function fitting. As a proof of concept, we measured the type and concentration of toluene and butanone multicomponent gas with a mean error of 6.34% and 4.32%, respectively. We analyzed the gas-sensing response mechanism, the theoretical support of the optimized parameters, and the causes of local errors. As a universal method, binary component gas detection was explored containing toluene, butanone, acetic anhydride, acetone, and diethyl ether.