1results about How to "Inhibit temperature drift" patented technology

A temperature compensation method for MEMS accelerometers based on fuzzy neural networks is presented, belonging to the field of MEMS accelerometer temperature compensation technology. This method addresses the problems of existing MEMS accelerometer temperature compensation methods, such as missing optimization of algorithm antecedent parameters, insufficient compensation accuracy, and poor real-time performance. The method includes: constructing an FCM-T-S fuzzy neural network module; optimizing the network antecedent parameters using the FCM algorithm; updating the network consequent parameters using gradient descent; and then deploying the trained FCM-T-S fuzzy neural network module into the MEMS accelerometer temperature compensation system. A data acquisition module synchronously acquires the temperature output voltage and acceleration output voltage of the MEMS accelerometer. The FCM-T-S fuzzy neural network module outputs the compensated true acceleration based on the temperature and acceleration output voltages. Results: High-precision temperature compensation for MEMS accelerometers is achieved, improving their output accuracy and temperature stability.