A Real Time Hardware Co-Simulation of Some Smart Features of Intelligent Sensor using ANFIS Architecture

Abstract

Due to time to market constraints, the application of Xilinx System Generator (XSG) and Matlab-Simulink which are consider as one of the rapid prototyping tools, has become increasingly important. In association with Field Programmable Gate Array (FPGA) devices, hardware co-simulation allows designers for evaluation, testing and validation of a new algorithm, a new component or a new prototype without damaging the actual system. This paper demonstrates a new method for the implementation of some smart features such as sensor fault detection and isolation and sensor drift compensation for intelligent sensors, using a real-time hardware co-simulation. This methodology helps to improve the design verification efficiency for such a system. In this paper, ANFIS (Adaptive Neuro Fuzzy Inference System) architecture has been used for detection and isolation of temperature sensor fault. The true estimated states are then compared with measured states and if the difference value exceeds the threshold value, the particular measurement is ignored and replaced by the true estimated state. Drift in resistance value is one of the important common failure modes for a nonlinear sensor such as thermistor. This is due to thermistor ages or parameter change which gives rise to inaccurate temperature measurements, affecting the performance of the system. FPGA based ANFIS compensator is used for carrying out the analysis of sensor fault or drift compensation in thermistor. The output of the thermistor is given as an input to the ANFIS compensator which then performs the corrective mapping of input signal to produce an output signal which is an estimation of the sensed signal. So, when sensor starts drifting, the drift is compensated by ANFIS compensator with barely affecting the performance of the thermistor. The presented results in this paper may be of great importance for design engineers.

Keywords: Hardware Co-Simulation, Sensor Intelligent Sensors, Fault Detection and Isolation, Sensor Drift Compensation, Thermistor, ANFIS, FPGA, XSG.