• Journal of Korea Society of Digital Industry and Information Management
• /
• v.17 no.3
• /
• pp.27-37
• /
• 2021

Highly informative heuristics in AI planning can help to a more efficient search a solutions. However, in general, to obtain informative heuristics from planning problem specifications requires a lot of computational effort. To address this problem, we propose a Partial Planning Graph(PPG) and Mixed Heuristics for solving planning problems more efficiently. The PPG is an improved graph to be applied to can find a partial heuristic value for each goal condition from the relaxed planning graph which is a means to get heuristics to solve planning problems. Mixed Heuristics using PPG requires size of each graph is relatively small and less computational effort as a partial plan generated for each goal condition compared to the existing planning graph. Mixed Heuristics using PPG can find partial interactions for each goal conditions in an effective way, then consider them in order to estimate the goal state heuristics. Therefore Mixed Heuristics can not only find interactions for each goal conditions more less computational effort, but also have high accuracy of heuristics than the existing max and additive heuristics. In this paper, we present the PPG and the algorithm for computing Mixed Heuristics, and then explain analysis to accuracy and the efficiency of the Mixed Heuristics.

• Industrial Engineering and Management Systems
• /
• v.9 no.3
• /
• pp.227-241
• /
• 2010

A memory module industry's supply chain usually consists of multiple manufacturing sites and multiple distribution centers. In order to fulfill the variety of demands from downstream customers, production planners need not only to decide the order allocation among multiple manufacturing sites but also to consider memory module industrial characteristics and supply chain constraints, such as multiple material substitution relationships, capacity, and transportation lead time, fluctuation of component purchasing prices and available supply quantities of critical materials (e.g., DRAM, chip), based on human experience. In this research, a directed graph-based supply network planning (DGSNP) model is developed for memory module industry. In addition to multi-site order allocation, the DGSNP model explicitly considers production planning for each manufacturing site, and purchasing planning from each supplier. First, the research formulates the supply network's structure and constraints in a directed-graph form. Then, a proposed genetic algorithm (GA) solves the matrix form which is transformed from the directed-graph model. Finally, the final matrix, with a calculated maximum profit, can be transformed back to a directed-graph based supply network plan as a reference for planners. The results of the illustrative experiments show that the DGSNP model, compared to current memory module industry practices, determines a convincing supply network planning solution, as measured by total profit.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.4
• /
• pp.709-714
• /
• 2023

A flexible planning system is an important element for the robot to perform various tasks. In this paper, we introduce an automated planning system architecture that can deal with the changing environment. PDDL is used for symbolic-based task planning, and a graph database is used for real-time environment information updates for automated PDDL generation. The proposed framework was verified through scenario-based experiments.

• The Journal of Korea Robotics Society
• /
• v.19 no.2
• /
• pp.196-202
• /
• 2024

The ability to understand given environments and plan a sequence of actions leading to goal state is crucial for personal service robots. With recent advancements in deep learning, numerous studies have proposed methods for state representation in planning. However, previous works lack explicit information about relationships between objects when the state observation is converted to a single visual embedding containing all state information. In this paper, we introduce graph-based state representation that incorporates both object and relationship features. To leverage these advantages in addressing the task planning problem, we propose a Graph Neural Network (GNN)-based subgoal prediction model. This model can extract rich information about object and their interconnected relationships from given state graph. Moreover, a search-based algorithm is integrated with pre-trained subgoal prediction model and state transition module to explore diverse states and find proper sequence of subgoals. The proposed method is trained with synthetic task dataset collected in simulation environment, demonstrating a higher success rate with fewer additional searches compared to baseline methods.

• The KIPS Transactions:PartB
• /
• v.18B no.1
• /
• pp.29-38
• /
• 2011

In order to derive domain-independent heuristics from the specification of a planning problem, it is required to relax the given problem and then solve the relaxed one. In this paper, we present a new planning graph, Merged Planning Graph(MPG), and GD heuristics for solving contingent planning problems with both uncertainty about the initial state and non-deterministic action effects. The merged planning graph is an extended one to be applied to the contingent planning problems from the relaxed planning graph, which is a common means to get effective heuristics for solving the classical planning problems. In order to get heuristics for solving the contingent planning problems with sensing actions and non-deterministic actions, the new graph utilizes additionally the effect-merge relaxations of these actions as well as the traditional delete relaxations. Proceeding parallel to the forward expansion of the merged planning graph, the computation of GD heuristic excludes the unnecessary redundant cost from estimating the minimal reachability cost to achieve the overall set of goals by analyzing interdependencies among goals or subgoals. Therefore, GD heuristics have the advantage that they usually require less computation time than the overlap heuristics, but are more informative than the max and the additive heuristics. In this paper, we explain the experimental analysis to show the accuracy and the search efficiency of the GD heuristics.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.13 no.2
• /
• pp.27-34
• /
• 2017

Highly informative admissible heuristics can help to conduct more efficient search for optimal solutions. However, in general, more informative ones of heuristics from planning problems requires lots of computational effort. To address this problem, we propose an Delete Relaxation based Action Costs-based Planning Graph(ACPG) and Action Costs-based Heuristics for solving optimal planning problems more efficiently. The ACPG is an extended one to be applied to can find action costs between subgoal & goal conditions from the Relaxed Planning Graph(RPG) which is a common means to get heuristics for solving the planning problems, Action Costs-based Heuristics utilizing ACPG can find action costs difference between subgoal & goal conditions in an effective way, and then consider them to estimate the goal distance. In this paper, we present the heuristics algorithm to compute Action Costs-based Heuristics, and then explain experimental analysis to investigate the efficiency and the accuracy of the Action Costs-based Heuristics.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.120.3-120
• /
• 2001

We propose a new method of path planning for cleaning robots. Path planning problem for cleaning robots is different from conventional path planning problems in which finding a collision-free trajectory from a start point to a goal point is focused. In the case of cleaning robots, however, a planned path should cover all area to be cleaned. To resolve this problem in a systematic way, we propose a method based on a graph model as follows: at first, partition a given map into proper regions, then transform a divided region to a vertex and a connectivity between regions to an edge of a graph. Finally, a region is divided into sub-regions so that the graph has a unary tree which is the simplest Hamilton path. The effectiveness of the proposed method is shown by computer simulation results.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.418-418
• /
• 2000

This paper proposes a path planning algorithm for mobile robot. To generate an optimal path and minimum time path for a mobile robot, we use the Genetic Algorithm(GA) and Visibility Graph. After finding a minimum-distance between start and goal point, the path is revised to find the minimum time path by path-smoothing algorithm. Simulation results show that the proposed algorithms are more effective.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.51 no.12
• /
• pp.577-582
• /
• 2002

The objective of this paper is to produce a weighted graph map for path-planning of an autonomous mobile robot(AMR) based on the measurements from a single ultrasonic sensor, which are acquired when the autonomous mobile robot explores unknown indoor circumstance. The AMR navigates in th unknown space by following the wall and gathers the range data using the ultrasonic sensor, from which the occupancy grid map is constructed by associating the range data with occupancy certainties. Then, the occupancy grid map is converted to a weighted graph map suing morphological image processing and thinning algorithms. the path- planning for autonomous navigation of a mobile robot can be carried out based on the occupancy grid map. These procedures are implemented and tested using an AMR, and primary results are presented in this paper.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.184-186
• /
• 2006

We propose a new method for shortest path planning of mobile robots. The topological information of the graph is obtained by thinning method to generate the collision-free path of robot. And the travelling path is determined through hierarchical planning stages. The first stage generates an initial path by use of Dijkstra's algorithm. The second stage then generates the final path by use of dynamic programming (DP). The DP adjusts the intial path to reduce the total travelling distance of robot. Simulation results are presented to verify the performance of the proposed method.