Guidelines to design a neural network as a feedforward controller for fast trajectory tracking of robotic arms
Bibliographic record
Abstract
Abstract Tracking fast and accurate trajectories of robotic arms can be important in applications involving large movements, velocities, and accelerations. This would require either an accurate dynamic model of the arm or an aggressive tracking with high-gain feedback. Concretely, it can be difficult to obtain the accurate model, due to nonlinearities and uncertainties. Current developments of 3D-printed and low-cost robotic arms accentuate this issue. Control architectures for high-speed trajectory tracking requiring no dynamic model were recently proposed. These consist in learning the dynamic response of a proportional derivative controller with a neural network (NN) as feedforward controller. However, no detail was provided to make the most of these architectures. This paper aims to provide guidelines for an optimal design of a neural network (NN) as feedforward controller for fast and accurate trajectory tracking of robotic arms. The subsequent objective is to compare 1. one NN per individual joint (INN’s method); and 2. one global NN (GNN method). The method compares these two architectures. Results are illustrated with two serial robotic arms of 3 and 5 degrees of freedom, simulated then in reality. The main results are as follows: The control architecture reduces the trajectory tracking errors (RMSE < 2°). The INN’s method can be used when the joints dynamics are decoupled and requires less data than GNN method to learn the dynamics. A table sums up the guidelines for design, in five main steps. Perspectives are to apply these guidelines to develop low-cost robotic arms and extend to micro-movements.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.007 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".