The mechanical properties of bamboo-wood composites (BWC) remain crucial for potential further employment. Nevertheless, the approach to detecting its mechanical properties is time-consuming and wasteful of resources. The goal to this research is to create an artificial neural network (ANN) model to predict the modulus of elasticity (MOE) of BWC using non-destructive testing methods of bending and longitudinal vibration to frequency, and to compare the results of the ANN with multiple linear regression (MLR). In comparison to the ANN model trained using spectrum parameters, the results reveal that the ANN model built using characteristic parameters delivers greater prediction performance. The ANN models had R2 and MAPE values of 0.98 and 3.187%, respectively. Using the findings of this work, the MOE of BWC can be calculated in a short amount of time with a low error rate, allowing researchers to better grasp the practicality of BWC in structural applications. Its artificial neural network model might be used to track the quality of BWC in the production process.
Read full abstract