This article introduces the principle of the glass curtain wall heating and cooling cycle constant temperature control test system in the building energy heating system and studies the constant temperature control strategy of its core cooling and heating circulation system. The author designed an adaptive fuzzy control algorithm with a self-learning function realize the thermal energy constant temperature control of the cooling and heating cycle experimental system and conducted the first model test. The test results show that the thermal energy constant temperature control strategy of the circulatory system and its adaptive fuzzy control algorithm have good control characteristics and effects. During the test, the thermal energy constant temperature control accuracy is within 0.4°C when the system is dynamically heated and cooled, and the temperature fluctuates when the constant temperature is maintained. Within 0.2°C, the test system fully meets the requirements of the American building energy heating AAMA501.5.98 glass curtain wall cycle test standard.