In order to address the problem of measurement inaccuracy in thrust measurement devices of micro-thrusters used in space gravitational wave detection caused by temperature variations during operation, a temperature-adaptive Smith fuzzy PID(proportional integral derivative) controller compensation scheme was proposed and implemented on an FPGA platform using a topdown multi-stage pipeline design approach and indirectly realised using the MATLAB fuzzy toolbox. The temperature delay problem was addressed by Smith prediction and compensation, and after verifying the algorithm model by simulation in Simulink, a fuzzy PID offline lookup table was established based on the Quartus Ⅱ development platform by using the FPGA's internal memory resources, which indirectly implements fuzzy reasoning and simplifies the complex reasoning process associated with traditional adaptive Smith fuzzy PID controllers. The simulation results demonstrate the efficacy of the proposed approach, substantiating the correctness of the functional design of the temperature adaptive control system. It is evident that the system effectively mitigates the complexity and slow response characteristics of traditional controllers, while concurrently attaining the stipulated design objectives.