Modeling oblique load carrying capacity of batter pile groups using neural network, random forest regression and M5 model tree

Experimental investigations on model vertical and batter pile group in uniform sands are presented. Mild steel piles in two different medium ofsand are used in this investigation. The tests are conducted on model steel pile installed in medium, and dense sand withL/d ratio is 18.75 and different batter angles of 0°, 10°, 20°, and 30°. These piles are constructed in sand and subjected to uplift loads of 60° inclination. It was found that the uplift capacity of vertical and batter piles under inclined pulls increased with increase in inclination of piles.it is also observed it a negative batter pile has greater uplift load than positive batter pile

Download Full-text

Knowledge-Based Prediction of Load-Carrying Capacity of RC Flat Slab through Neural Network and FEM

Mathematical Problems in Engineering ◽

10.1155/2021/4528945 ◽

2021 ◽

Vol 2021 ◽

pp. 1-18

Author(s):

Afaq Ahmad ◽

Muhammad Usman Arshid ◽

Toqeer Mahmood ◽

Naveed Ahmad ◽

Abdul Waheed ◽

...

Keyword(s):

Neural Network ◽

Carrying Capacity ◽

Research Work ◽

Compressive Force ◽

Critical Parameters ◽

Load Carrying Capacity ◽

Concentrated Load ◽

Flat Slab ◽

Load Carrying ◽

Input Parameters

The present research work aims to compare the results for predicting the ultimate response of Reinforced Concrete (RC) members using Current Design Codes (CDCs), an alternative method based on the Compressive Force Path (CFP) method, and Artificial Neural Network (ANN). For this purpose, the database of 145 samples of RC Flat Slab with the simple supported condition under concentrated load is developed from the latest published work. All the cases studied were Square Concrete Slabs (SCS). The critical parameters used as input for the study were column dimension, cs, depth of the slab, ds, shear span ratio, a v s / d , longitudinal percentage steel ratio, ρls, yield strength of longitudinal steel, fyls, the compressive strength of concrete, fcs, and ultimate load-carrying capacity, Vus. Seven ANN models were trained using different combinations of input parameters and different points of hidden neurons with different activation functions. The results exhibited that SCS-4 was the most optimized ANN model, having the maximum value of R (89%) with the least values of MSE (0.62%) and MAE (6.2%). It did not only reduce the error but also predicted accurate results with the least quantity of input parameters. The predictions obtained from the studied models (i.e., CDCs, CFP, and ANN) exhibited that results obtained using the ANNs model correlated well with the experimental data. Furthermore, the FEM results for the selected cases show the closer result to the ANN predictions.

Download Full-text

Determining maximum load-carrying capacity of robots using adaptive robust neural controller

Robotica ◽

10.1017/s0263574710000081 ◽

2010 ◽

Vol 28 (7) ◽

pp. 1083-1093 ◽

Cited By ~ 2

Author(s):

M. H. Korayem ◽

A. Alamdari ◽

R. Haghighi ◽

A. H. Korayem

Keyword(s):

Neural Network ◽

Carrying Capacity ◽

Industrial Robot ◽

Maximum Load ◽

Load Carrying Capacity ◽

Back Propagation Algorithm ◽

Robust Controller ◽

Linearization Methods ◽

Load Carrying ◽

Articulated Robots

SUMMARYIn this paper, the combination of neural network (NN), proportional derivative (PD), and robust controller are used for determining the maximum load-carrying capacity (MLCC) of articulated robots, subject to both actuator and end-effector deflection constraints. The proposed technique is then applied to articulated robots, and MLCC is obtained for a given trajectory. In the practical simulations, it's impossible to determine the parameters of robot model exactly, so the trajectory tracking performance of the proportional integral derivative (PID) and computed torque methods significantly decrease. The PD control of robot has major problem, it cannot guarantee zero steady state error. For this reason, the NN controller with PD and robust controller are used. The multilayer neural network is also used to compensate gravity and friction effects. By using Lyapunov Direct Method it is shown that the stability of closed loop system would be guaranteed, if the weights of multilayer had certain learning rules. Standard back propagation algorithm is used as a learning algorithm to update the connection weights of the NN controller. The simulation results of the proposed adaptive robust neural network (ARNN) controller are compred with sliding mode and feedback linearization methods for flexible joint robot, and compared with open loop controller for 3D industrial robot. The obtained results assured the robustness and improvement in MLCC in the presence of uncertainties in dynamic model of the robot arm and external disturbances. In fact, adaptive robust NN controller suppresses disturbances accurately and achieves very small errors between commanded and actual trajectories.

Download Full-text