Temporal Planning with Semantic Attachment of Non-Linear Monotonic Continuous Behaviours
Authors: Josef Bajada, Maria Fox, Derek Long
IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | A new planning system implementing this algorithm was developed and evaluated. Results show that the addition of this algorithm to the planning process can enable it to solve a broader set of planning problems. ...A PDDL 2.1 temporal planner, u NICOrn, was developed in order to evaluate this algorithm... This planner was evaluated using two domains with non-linear continuous effects. Results show the effectiveness of this algorithm in solving this class of problems and the impact of decreasing the error tolerance on its overall performance. ...6 Empirical Evaluation The non-linear convergence algorithm was integrated with the planning framework described in Sections 3 and 4. The implemented planner, u NICOrn, performs a breadth-first search (BFS)... The system was evaluated using two domains... Tests were performed using an Intel R Core TM i7-3770 CPU @ 3.40GHz. ...Figure 3 shows the time taken to produce a plan for this domain, for up to ten tanks. ...Table 1 compares the performance of u NICOrn (with 0.1 error tolerance) with that of UPMurphi (with 1.0 time discretisation) using a maximum of 3GB memory on the same setup. |
| Researcher Affiliation | Academia | Josef Bajada and Maria Fox and Derek Long Department of Informatics, King s College London London WC2R 2LS, United Kingdom {josef.bajada, maria.fox, derek.long}@kcl.ac.uk |
| Pseudocode | Yes | Algorithm 1: Non-linear Iterative Convergence |
| Open Source Code | No | The paper mentions developing a new planning system called 'u NICOrn' but does not state that its source code is publicly available. |
| Open Datasets | Yes | The system was evaluated using two domains, the Tanks domain (described below), and the Car domain [Fox, 2006]. |
| Dataset Splits | No | The paper does not specify training, validation, or test dataset splits (e.g., percentages or sample counts). |
| Hardware Specification | Yes | Tests were performed using an Intel R Core TM i7-3770 CPU @ 3.40GHz. |
| Software Dependencies | No | The paper mentions various systems and languages like 'PDDL 2.1', 'u NICOrn', 'TM-LPSAT', 'COLIN', 'POPF', 'OPTIC', 'UPMurphi', and 'VAL' but does not provide specific version numbers for these or other software dependencies. |
| Experiment Setup | Yes | Each semantically attached function can have a dedicated margin of error, within which each non-linear continuous effect is required to be estimated. The system also enables the user to optionally specify this as a parameter within the problem definition. ...Listing 4 shows the output plan for the Tanks problem with 3 tanks, with error tolerance em = 0.0001 and ε = 0.001. ...Each problem was also executed with a range of error tolerance values to analyse the impact of the increase in the number of iterations needed to converge. ...Table 1 compares the performance of u NICOrn (with 0.1 error tolerance) with that of UPMurphi (with 1.0 time discretisation) using a maximum of 3GB memory on the same setup. |