FPGA Embedded Signal Conditioning System Based on Fuzzy Logic for Temperature Measurement

Temperature is an important physical variable that is measured and controlled in many manufacturing stages in industrial processes. Commonly, a temperature measuring digital system requires additional signal conditioning circuitry, which represents a restriction for small area, weight, and power consumption applications. An FPGA embedded signal conditioning system based on fuzzy logic for temperature measurement is proposed in this paper. In this way external signal conditioning hardware is not needed for the temperature measurement. Particularly, a Takagi–Sugeno-Kang fuzzy inference system was employed. The temperature measurement system was implemented on a Spartan 6 FPGA using an LM35 temperature sensor. The sampled temperature value is real time monitored and displayed on a PC screen through RS-232 serial communication protocol. The measurement system using the proposed fuzzy logic signal conditioning was experimentally tested within a temperature values range, which was compared with the obtained values when it was not used. According to the statistical analysis of the obtained results, an average error reduction from 305.15% to 2% and a variance error reduction from 1851.3% up to 2.64% were achieved. The described fuzzy logic system can be extended to conditioning other analog signals by just adjusting the membership functions coefficients, which is left as future work by the authors.